|
Enough of these overdone pseudo-HDRs
 |
They're not real HDRs and they're almost always overdone and not what HDR was intended to do.
Look I have nothing against what you consider 'artistic expression'. Just start another group for it or join one of the existing groups already on Flickr:
www.flickr.com/groups/fhdr/
www.flickr.com/groups/raw2hdr/ (actually a misnomer because a single RAW could not technically result in an HDR image)
And please stop calling it HDR. The generally agreed term for a single tone-mapped RAW is pseudo-HDR.
Originally posted at 5:18PM, 12 October 2008 PST
(
permalink
)
waynewhuang edited this topic 2 months ago.
|
 |
Is your beef with the look of Psuedo-HDRs, or because they are incorrectly called HDRs?
I think both these topics have been discussed ad infinitum.
Personally I think a lot of the confusion would be cleared up if there was a clearer distinction between the process of producing a high-dynamic-range photo and the process of tone-mapping. I think there are a great number of people who mistakenly think tone-mapping is HDR. Hence some people see an ordinary JPG processed with the Lucisart filter and immediately think it is HDR.
This also explains why people who use PS to produce an HDR are often dissapointed with the results: PS doesn't use the Photomatix tone-mapping algorithm that people often find so appealing.
Originally posted 2 months ago.
(
permalink
)
chidginess edited this topic 2 months ago.
|
 |
I've never seen a real HDR image on a computer monitor, here on Flickr or anywhere else. They're usually just 8-bit JPG's, sometimes a 16-bit image file (which I can't actually see on my monitor either)
Originally posted 2 months ago.
(
permalink
)
skinjester edited this topic 2 months ago.
|
 |
Dont feed the trolls
Posted 2 months ago.
(
permalink
)
|
 |
lol
Posted 2 months ago.
(
permalink
)
|
 |
HDR and tonemapping seems to have touched a nerve with some. The post above essentially is critical of many of the tonemapped 'HDR' photos that are making their way onto Flickr. Semantics aside, I'd have to say that I disagree with the poster. I think what we are seeing is the development of a new way of expression. The 'HDR/tonemapping' tools (whatever you want to call it) are responsible for quite a few of these 'baked' images showing up. I love them. I don't really care if it breaks somebody's preconceived notions of what photo rules are supposed to be. Its all about expression. This is just another form, no?
Posted 2 months ago.
(
permalink
)
|
 |
So, Wayne, what WAS HDR 'intended' to do, exactly?
I'd be interested in hearing what you think....
Posted 2 months ago.
(
permalink
)
|
 |
Your photo is YOUR artwork. While HDR may not be natural looking, I'd hardly call some abstract art natural looking. Personally the car image above is pretty cool!
Posted 2 months ago.
(
permalink
)
|
 |
i second kcrusher... seriously... what is the intent of HDR? because last time i heard there wasn't some photowizard who invented HDR and decides what is right and what is wrong... seriously most people who are posting on here are just having fun who are you to say what the intent of HDR is... some people have the intent to make them all crazy and overprocessed that's there choice... HDR is just a process just like any other photographic process... and in this crazy new modern age with these fancy computer matrix devices people can get a bit carried away... but i'm sure when HDR was used in film photography people got carried away as well... and i'm sure some people cheated to achieve that same effect... so for you to say that someone should join another group or stop posting on here is pretty lame dude... just let people have fun and quit being such a bummer.
Posted 2 months ago.
(
permalink
)
|
 |
lmfao, way to make a stand!
YOU TELL EM!
rinse, repeat *inf
Posted 2 months ago.
(
permalink
)
|
 |
True enough Troubledog!
Posted 2 months ago.
(
permalink
)
|
 |
Hot topic. Well, personally I wouldn't be as harsh as wayne but I wouldn't trash him either. Maybe there is some merit in differentiating between images that are derived from a number of bracketed shots that have been merged into a HDR file and then tonemapped and images derived from a single tonemapped shot.
For me the beauty of "real HDR images" that are derived from merging more than two bracketed photos is that the right tonemapping can reveal details and effects that are simply not available from single shots. The tonemapping process renders us endless opportunities for new effects and expressions from natural looking to comic book style; everyone can do what he likes best.
For me a "real HDR image" should reveal clean details from shadows AND highlights that cannot be captured with a single shot. Sure, many of the photos shown here don't need to be taken from multiple shots simply because the dynamic range in the scene is small enough to fit into a single shot.
The tonemapping is just another form of postprocessing - some like the effects, some don't. If a real HDR image is tonemapped it can reveal much more than an image produced from a single file - and if a single RAW is tonemapped it will be just another heavily postprocessed image that some might call "pseudo HDR".
It makes sense to me to have two threads: One thread for tonemapped images derived from merging multiple bracketted shots ("real HDR") and another thread for tonemapped single shots ("pseudo HDR").
Originally posted 2 months ago.
(
permalink
)
Bernd Mechsner edited this topic 2 months ago.
|
 |
:)
Originally posted 2 months ago.
(
permalink
)
Unknown Revelation edited this topic 2 months ago.
|
 |
I'll be honest when I first got my my camera (and consequently joined flickr) I loved these over the top tonemapped hdr's but now i've totally lost interest in them. I find the whole process a bit too formulaic and because of that not an extremely creative form of photography, but thats just my opinion, and at the end of the day they are fun and can have a stunning impact and wow factor if done well so each to their own. Surely if you are a hobby photographer (like 90% of people on here) you should just take/make pictures you enjoy taking/making and have fun with it, so I say leave the OTT hdr's alone because all pictures have got their place on flickr.
:-)
Posted 2 months ago.
(
permalink
)
|
|
@Unknown Revelation
Good example of a tonemapped image made from 9 merged shots. A single shot would never reveal clear details inside the house no matter what tonemapping or other treatment you apply. This image makes a huge dynamic range visible to us. We see in one image the details of a much higher dynamic range than a modern imaging sensor can capture into a single shot - and that's what HDR is all about. The tonemapping compresses the high dynamic range of multiple merged shots into a range that can be displayed on our monitors and thus the normally "hidden" data from shadows and highlghts are made visible. The cartoon-like look and the oversaturated colors are just a result of the adjustments of the tonemapper according to the taste of the photographer.
Posted 2 months ago.
(
permalink
)
|
 |
Yes, this is a good example of what HDR can do. But, it's the "cartoon-like look and the oversaturated colors" that makes the image just another over the top image. It's too bad.
Posted 2 months ago.
(
permalink
)
|
|
@eminer
You are right. I share your taste about the "cartoon-like look and the oversaturated colors" but I tried to keep my remarks neutral...
Personally I love photos that can reveal a magnificient dynamic range but that are not tonemapped like cartoons - but that's just my personal preference.
Posted 2 months ago.
(
permalink
)
|
 |
waynewhuang,
Do you know when you shoot in RAW the sensor captures more info in one image than can be processed by standard monitors?
When you develope a RAW you are only taking part of that data and clipping the rest thus losing highlight and shadow info and detail. It is possible to develope a RAW image with several Exp values to capture all of what the sensor captures and process using HDR to gain lost detail.. That my friend is technically providing a higher dynamic range than if you just develope a single .tiff or .jpg which is exactly what HDR was developed to do.
While I agree multiple exposures will give you a higher level of detail, I strongly disagree that nothing can be gained from HDR from a single RAW..
There are various books on this subject and plenty of facts and information to back this up.
Posted 2 months ago.
(
permalink
)
|
|
maven17 - Could you please point me to info that backs up this statement "Do you know when you shoot in RAW the sensor captures more info in one image than can be processed by standard monitors?" Because I don't beleive that's true. A RAW image is contains the same exposure info as any other format does.
I'm not sure what you mean by this - "I strongly disagree that nothing can be gained from HDR from a single RAW.. "
Are you saying that a single RAW image contains the same ammount of info as an HDR image?
Posted 2 months ago.
(
permalink
)
|
|
Eminer1254,
I'm not saying RAW contains the same ammount of info as HDR, but it does contain more than any single image developed from that RAW. That is why you need a program or plug-in to process camera RAW images, because they are just that..raw data.
When you develope a RAW to .tiff or .jpg or whatever format you choose, you are not getting all the data available, just what had been processed based on your preferences set by whatever progam you use.
Here is one helpful link to understand more of what's going on when shooting RAW.
www.luminous-landscape.com/tutorials/understanding-series...
And here's a link to what I feel is the best book on the subject...
www.amazon.com/HDRI-Handbook-Dynamic-Imaging-Photographer...
I provided all this not so much to add to the argument, but more because I feel some people are just are not aware of all the tech details.. Hope it helps! :-)
Posted 2 months ago.
(
permalink
)
|
 |
What he ( ) said!
Posted 2 months ago.
(
permalink
)
|
|
I have never, ever seen a cartoon that looks anything like that. But thanks for the feedback. I only posted it to annoy the OP.
Posted 2 months ago.
(
permalink
)
|
|
@maven
You are looking at imaging from a little bit of a different aspect. I certainly agree that a 12bit or 14bit RAW image contains more data than what can be displayed on a monitor - and it certainly contains much more data than any image derived from an 8bit format file. However whatever you do with a single Raw file is just postprocessing of a single image to make some of your normally hidden image information more visible. Whatever you get from massaging curves and highlights and shadows and contrasts or from tonemapping is still not made of more data than what the sensor of your camera can handle. No postprocessing or tonemapping process can get information out of fully blown highlights or clipped shadows - you cannot get information from where there are no information.
Let's have a look at the facts you will find documented in the literature: The best image sensor can capture a usable dynamic range of 9 to a bit more than10 f-stops. The best monitors can display a range from 8 to 9 f-stops max. Single view human vision without dark adaptation has a dynamic range of 13-14 f-stops and with dark adaptation the human eye can have a range of up to 27 f-stops. No matter what treatment or postprocessing or tonemapping you will do, an image derived from a single RAW file will still reveal at best a range of about 9 to 10 f-stops. And that is much less than what the human eye can see.
If you want to capture a magnificient scene with a higher dynamic range than 9-10 f-stops, then there is no other way than to bracket shots of different exposure values, merge them into a HDR file and tonemap them into a visible picture displaying a range of 8-9 f-stops. And for me that's what HDR imaging is all about: that is getting a single image that can reveal details from the entire visible dynamic range and not just from the range the image sensor can capture. Unfortunately this cannot be done with treating a single RAW file.
*edit: Here an excellent source for a good description not only of the dynamic ranges relevant to cameras, monitors, humans and various scenarios: "Mastering High Dynamic Range Photography", written by Michael Freeman
Originally posted 2 months ago.
(
permalink
)
Bernd Mechsner edited this topic 2 months ago.
|
|
Bernd Mechsner
I agree with you whole heartedly about the process and to get a truly closer to what the eye sees you need multiple exposures..
Considering true HDR can not be displayed on standard monitors and everything here tagged with "HDR" is saved to a lossy 8 bit format, someone declaring true and non true HDR is kind of pointless anyway.
My main point to waynewhuang was that you can in fact get a larger dynamic range from a single RAW that just processing it as a single image. Granted it is not as detailed as a multi-exposure and it does not capture the full range.. But it is a Higher Dynamic Range than a simple single .jpg so to some degree it is HDR because you are getting details that would otherwise be clipped without this process.
But to each his or her own.. It's an evolving tech and it truly is being redefined daily. I just hate to see people label and segregate because it doesn't fit a specific view of how to do it..
To me anything that achieves more than a standard image in detail is HDR
Posted 2 months ago.
(
permalink
)
|
 |
@maven17 -- a *very* common misconception is the difference between the brightness/luminance levels captured in a RAW file and true floating point luminance values. They are NOT the same thing.
Warning: Long explanation ahead...
When your camera sensor "sees" an image, what it is recording is the amount of light hitting it. We all know this.
When the processor records the data from the sensor, it does 2 things: 1) it records the brightness of the light hitting the photosite, and 2) it assigns that light a color value (either Red, Green or Blue).
As far as the photosite is concerned, no light hitting it would be black, or a value of 0, and the maximum light it can record would be white (or, when a color is assigned, the full brightness of that color - "red" for example).
Where the bit depth of the RAW file comes into play is in the number of steps in between black and white. An 8-bit file has only 256 possible values the processor can assign to what the photosites have captured for the Red channel, 256 it can assign to the Green channel, and 256 it can assign to the Blue channel. So no light, or black would have a value of 0, while "maximum" light would have a value of 255. A 12-bit can record values from 0 to 4,095 per channel, while a 14-bit can record from 0 to 16,383 per channel.
Note: This does NOT translate into "more" colors or "more" dynamic range. Black is still black and white is still white. What the added bit-depth does is increase the accuracy by giving more "steps" between black and white to get the color right.
So what you get out of a RAW file is the potential for *more accurate* colors.
Now, here's where it gets tricky. As I said above, each photosite only assigns ONE color value when it captures an image. So a given photosite might assign its value to Red, while the photosite right next to it might assign itself to Blue. HOWEVER, the image you look at doesn't just look like a bunch of red, green and blue dots. That's because the RAW file has to be converted -- a process called demosaicing. This blends the RGB values into the full color spectrum you expect to see in your image.
When you shoot in JPEG mode, you essentially let the camera make the decision over the best way to blend the RGB values, and it "locks" the image into the JPEG format.
When you shoot in RAW, the camera creates a "basic" demosaiced image based on your camera settings, but *keeps* the RAW values, allowing you to adjust the blending in your RAW editor.
Adjusting your RAW files is essentially just playing with the brightness of the RGB values -- a little more red here, a little less blue there, etc.
But the important thing to keep in mind is that the adjustments you are making are all software-based exposure compensation (EC).
Software EC is very different from in-camera EC. Your camera's exposure compensation essentially shifts the entire range up or down, capturing cleaner data. Software, on the other hand, can't create data that isn't there -- whatever it captured for black and whatever it captured for white are the absolute limits of the captured data. Whatever goes beyond those limits is clipped and replaced by noise -- and *ANY* adjustment up or down to the RAW file creates some amount of clipping. The question is just how visible that clipping is -- even though data is lost, a stop up or down tends to still *look* pretty decent. Two stops and you start to really notice it (whereas in-camera EC by two stops may be either over or under exposed, but most likely also is not clipping the data).
To add one more level of complexity to the mix, the brightness / luminance values recorded in a RAW file can ONLY be integers. So brightness values of 5.1 and 5.9 are, as far as the RAW is concerned, each only 5.
What this means is that RAW files are great at capturing the values as set by the camera (which is essentially recorded as the "base" exposure). However, because the RAW file only deals with integers, any adjustment up or down is prone to errors -- and the further you try to adjust, the bigger the errors are. For example, when adjusting up from, say, a value of 5, you might be able to go to 6 or 7, but what's really most accurate is 6.5. On the small scale, this is no big deal, but errors grow the further you try to adjust from the "base."
In other words, a RAW file adjusted to be a stop brighter might look pretty good to your eye, but is not mathematically as accurate as a shot with in-camera EC set to a stop brighter.
A 32-bit HDRi file is different from other image files in that it records true floating point numbers -- it isn't limited by integers. HOWEVER, if the errors are in the file to begin with, they won't magically be fixed in the conversion to a 32-bit file.
Very long story short:
1. The dynamic range captured in a RAW is not the same as the dynamic range found in separate, multiple exposures.
2. Can you tone map a single RAW or fake exposures from a single RAW and tone map those and get decent looking results? Sure. But you don't have a high dynamic range representation, and you would *still* get a more accurate image from true multiple exposures (even if you're talking only a 1-stop up and down range).
Posted 2 months ago.
(
permalink
)
|
 |
For those who don't know this already...
HDR was originally created for Architects, so that they could render 3D environments scenes with real world lighting values, cause previous to this Architects only worried about the design of the building but now could include lighting to be an integral part of the design process. Greg Ward.
This was later pushed further by Paul Debevec who showed how taking separate exposures could be merged together to produce a Radiance map which would then be used to light scenes up in 3D environments. This technology was called IBL or Image Based Lighting. If look at how lighting was done previous to HDR you would understand what difference this made.
Here is a link to the brief history of HDR
www.flickr.com/groups/artizenhdr/discuss/72157594280410378/
Posted 2 months ago.
(
permalink
)
|
|
Seems I've engendered a bit of hostility, totally not my intention. I guess the group has just become a little convoluted with non-HDR imagery and total misrepresentations of it.
It's just frustrating to have to browse through more non-HDR images just to see real, well-done HDR images that inspire new ideas and ways of using HDR.
I've created a new group, Naturalistic HDR Photography, that I hope some of you purists here will join.
www.flickr.com/groups/879050@N24/
Originally posted 2 months ago.
(
permalink
)
waynewhuang edited this topic 2 months ago.
|
 |
You'd have to understand sensitometry and densitometry to know fully how HDR effects tonal range. HDR is not new....it's been around since film days and before home computers. It took all day to produce a single usable image with reduced shadows and highlights by using a series of masking inter-negatives sandwiched together.
One day, I'm going to use a Kodak 21-step tablet to illustrate exactly what happens.
Personally, I like the look of some single-image psuedo-HDR's....but not all of them. Many people overcook them and up the color saturation thinking they can make it artsy, when in fact there's few people who this appeals to.
Posted 2 months ago.
(
permalink
)
|
 |
Wayne, this is something of a hot topic, and for some weird reason it always brings out the worst in people.
Every time I've seen the subject raised, it always degrades into abuse, name calling and petty bickering.
It's just photography people...Chill, relax and stop taking life so seriously.
@ Wayne, not everyone will agree with your philosophy, so trying to convince them is a waste of effort
@ Everyone else, Wayne is just as entitled to voice his opinion as you are, and with a topic as subjective as photography and HDR, his opinion is just as valid as yours or mine.
For what it's worth, I generally agree with Waynes original post...However...
HDR is an artistic technique, so there's infinite room for different interpretations of the technique and effect.
For example, Pablo Picasso was taught to draw and paint in the conventional way, creating pretty images of landscapes, portraits and still life's.
But those are not the pictures he is remembered for.
He's remembered because at some point in his career, he decided to break those rules and develop a unique and wacky style all of his own. It may not have been to everyones taste, but over time he has become accepted as one of the masters, and his style just became another evolutionary branch of art history.
Similarly, would anybody remember Jimmi Hendrix if he had kept on strumming tame folk songs on an acoustic guitar?
No.
He broke the moulds, and created a whole new genre of acid fuelled blues/rock fusion.
As a result, he revolutionised the music world forever.
He didn't kill off the existing styles.
People still continue to play folk, country, blues, classical, motown and funk.
He merely added a new style of music to the ever growing list that exists, making the music world a more diverse place.
My point here is that just because you and I personally don't like a particular style of tonemapping, we should not try to stop it or censor it.
Even if only one person likes it, then it becomes a valid artform.
If you try to stop a new style or technique, you stifle creativity.
If you stifle creativity, you effectively stop artistic evolution dead in it's tracks.
That's definitely NOT a good thing.
Originally posted 2 months ago.
(
permalink
)
Sam Knox edited this topic 2 months ago.
|
 |
Sam Knox, agree.
"Rules are like hearts - They`re meant to be broken."
(altought the same thing can be said for promises) :D
Posted 2 months ago.
(
permalink
)
|
|
@Sam Knox -- I agree with the spirit of what you wrote, and I am firmly in the "people should create the images they want and I not everyone has to like it" camp.
However, and I don't intend this to be mean toward anyone, comparing super saturated, contrast-less, tone mapping to the works of Picasso and Hendrix is a rather large stretch. Those two artists were innovators. There really isn't anything innovative about over-done tone mapping, at least not anymore. In fact, depending on the software used, creating images that don't look like they've been baked in the oven 20 minutes beyond "well done" is what takes some real creativity.
A better comparison might be the speed oil painting techniques of the late, great Bob Ross. Sure, many were snobby toward his "happy little trees" and generic landscapes, but for quite a number of people, his techniques became their outlet for their own artistic expression. ; )
Posted 2 months ago.
(
permalink
)
|
|
wayne i agree you have the right to have your own opinion and i would never judge you for it. I only have issue when someone says that HDR is not intended for a very specific purpose. It would be like saying i use my Cameras as paper weights so you can't tell what it is "intended for". HDR has a very specific origin with very specific requirements that has changed the movie and architectural industries.
That being said it doesn't mean ppl can't use the HDR technology to express and artistic idea. All this not a real HDR not a natural HDR is all stupidity and Much ado about Nothing. Use it for what you want and have fun, but when you are making a movie with CG you will then know what HDR not tone mapped HDR was really for.
Posted 2 months ago.
(
permalink
)
|
 |
For those of you who think that you can't 'add' information that isn't there in a single raw photo.
"Highlight recovery
DxO Lighting and Highlight Recovery
A common problem in digital photography is clipped highlights, when pixels have maximum values in one or more of the colour channels (red, green or blue). DxO’s Highlight Recovery feature can correct this when working with RAW images. Provided that only one channel is clipped, the software can restore lost tonal information. It does this by analysing the ratio of R, G and B colours in adjacent unclipped pixels and making local gamma adjustments to assign new values to the affected area. The result is that natural-looking highlight detail can be retrieved from overexposed areas." -from DxO's website.
Granted it's not going to be like using multiple exposures, but to me that's adding information that isn't in the initial exposure.
Here's an example of above quote in use.
Initial exposure.
Final result.
Originally posted 2 months ago.
(
permalink
)
Emteehed edited this topic 2 months ago.
|
 |
Emteehed, Clever stuff indeed.
Posted 2 months ago.
(
permalink
)
|
|
@Emteehed -- The point isn't just adding information. Rather it's about real information from the scene.
What you've posted is a very good example of what can be done to generate a more usable image from a poor exposure. However, interpolating data isn't the same thing as recovering *actual* data. Just as interpolating pixels to create a larger image has limitations, there is going to be a point of diminishing returns with this process.
I'm not trying to dismiss the results, because I think it is a good recovery tool. But at the end of the day, the software is still not recovering actual data *from the scene* that was missing in the exposure. It is only making its best guess about what that information *might* have been.
So, sure, it is adding information that wasn't there, but because it is interpolated information, it's a gamble about the accuracy. The very quote you posted said it best -- "natural-looking" highlights, versus the ability to recover "real" highlights.
So if it makes you happy, the point can be clarified to "you can't invent information *that was actually in the scene* if it wasn't captured in the original exposure." ; )
Posted 2 months ago.
(
permalink
)
|
|
I agree with you Zuiun partially. As I previously said, it is not like taking multiple exposures.
But that really isn't a poor exposure. Considering the amount of contrast in the scene, that is how I had to expose it to come up with the final result. It was planned and intended to be exposed that way purposely for the intentions of the final result.
What you consider to be poor was actually putting the most use to the sensor as I could with what knowledge I have of how DxO works.
Maybe you don't like the image itself and I can agree that it's not an excellent image, but for the purpose of what I was saying, it was an excellent example. Please don't try to downplay my photographic skills. I may not be a paid professional, but I can assure you that I knew what the heck I was doing in getting the exposure I needed.
Posted 2 months ago.
(
permalink
)
|
|
@Emteehed -- Whoa.... where did I try to "downplay" your photographic skills? Or even say that I didn't "like" the image? Now you're trying to invent intent -- see how much trouble interpolation can cause? ; )
This discussion -- i.e., a discussion about HDR and the merits of multiple exposures -- is about DETAIL in the image. From that perspective, it *is* a poor exposure. That has nothing to do with your skills or your intent for the image.
Bottom line, your image lacks details -- some unrecoverable -- in both the highlights and shadows.
That fact that you got the image you intended has NOTHING to do with the detail in the image.
I'm quite fond of this image of mine, and it's the exact look I was after:
But from a detail standpoint, it's VERY poor and even if I wanted to, nothing can recover the lost detail.
Posted 2 months ago.
(
permalink
)
|
|
Calling it a 'poor exposure' is exactly what I would call 'downplaying my photographic skills'. Another way of saying, 'maybe I should learn to expose my scenes correctly'.
Here is a closer look. Now if the scratches and gunk on the window is a sign of not recovering 'detail', then maybe you are right and I am wrong. But as it appears in the initial exposure, those details are not there.
As far as the discussion topic, HDR and multiple exposure, the final result was from taking a single RAW and breaking it down into several images 1 EV apart instead of just tone mapping the single RAW.
1. Do the stand alone HDR programs perform recovery such as DxO's highlight recovery function?
2. I usually come out with better images breaking the RAW into several images rather then just tone mapping the RAW. Partially because I will purposely clip a channel in order to try and not lose so much detail in the deeper shadows. Hence breaking the RAW down highlights have to be recovered first before merging.
3. In having separated the RAW into several images, I then had to merge them back, thereby becoming a 32-bit floating point file.
Posted 2 months ago.
(
permalink
)
|
|
@Emteehead
Cool example!
You show a nice effect but in my eyes that has nothing to do with HDR. The software in your example is just working with the data that the camera sensor could write into the RAW file. Since the best sensors can only capture 9-10 f-stops it is still a limited dynamic range. The beautiful thing of merging multiple bracketed shots into a 32bit HDR file is that then your software has got real data to work with and can reveal real details from a range that is well above the usual 9-10 f-stops. How one does handle the data from the HDR file depends on the software or tonemapper one is using - and for this there are endless possibilities and that's where one can put most creativity into his shot..
Posted 2 months ago.
(
permalink
)
|
|
The term 'pseudo-HDR' is in the title of the topic. And I did say earlier, that it's not the same as taking multiple exposures.
Being in a plane going something like 300 MPH or so, it seems multiple exposures to be a little difficult.
Posted 2 months ago.
(
permalink
)
|
|
@Emteehead
Sorry, I was writing when you were posting your last message.
What you are doing is using software and technique to make more details visible. These are details that are already captured by your sensor but conventional software is not subtle enough to make them visible.
HOWEVER this is not added dynamic range. These details are nice but they don't add dynamic range or f-stops to your image. So again it has nothing to do with HDR. You are using a treatment like it is used for HDR images but it has nothing to do with higher dynamic range.
If you look at the literature published about HDR then it seems clear that HDR photography is all about adding details from a higher dynamic range than what can be captured by a sensor in a single shot. HDR like treatment like you do is nice and it reveals some more details within the limited dynamic range of your sensor BUT again it doesn't reveal anything beyond the 9-10 f-stops dynamic range of your camera sensor.
Posted 2 months ago.
(
permalink
)
|
|
NO SHIT!!! It's not like taking a photo like this.
The point of what I was saying was against people saying that you can't add what isn't in the single RAW. And to a certain extent, you can. The thing about it is, do the HDR programs recover highlight details the same way? In my experience, no they don't. Photomatix wasn't able to produce the same image from just tone mapping it. So basically i have to break the RAW into several exposures and merge them back into an HDR file. <<<< This is where HDR comes into play. Granted it's still not a TRUE HDR with the huge dynamic range in contrast.
Originally posted 2 months ago.
(
permalink
)
Emteehed edited this topic 2 months ago.
|
|
@emteehed -- you wrote:
Calling it a 'poor exposure' is exactly what I would call 'downplaying my photographic skills'. Another way of saying, 'maybe I should learn to expose my scenes correctly'.
Um... no. "Expose your scene correctly" isn't what I wrote at all. But whatever, dude. If you want to take it personally, that's your own issue to deal with.
Regarding your photo, the large size version of the original image is unavailable. Furthermore, no one here has access to your original RAW file, so there is no way to know if DxO was the *only* or even the best way to recover those dirt details on the window. The key is as it "appeared" in the original exposure. Obviously those dirt details were recorded in *some* fashion.
Look, there is NO dispute whatsoever that a RAW file records more recoverable data than is available in an 8-bit image. Were that not the case, there would be no point in shooting RAW. But the issue is HDR and the detail present in capturing true multiple exposures, versus thinking that a single file can retain the same level of detail.
As you seem to be in agreement with, it cannot. Trying to recover the SAME level of detail by altering the exposures of a RAW will not yield the same accuracy as taking genuine multiple exposures. Period.
That isn't to say that single file conversions can't look good -- a point I made waaaaay up there in my first post in this thread.
Also, merely making multiple images into a 32-bit floating point file isn't, in itself, meaningful. You can merge the same 3 exposures into a 32-bit floating point file and that doesn't mean you have any more data there. The file format itself isn't magic. It merely has the POTENTIAL to store exponentially more information than linear files.
Posted 2 months ago.
(
permalink
)
|
|
@ Emteehed -- you wrote:
Being in a plane going something like 300 MPH or so, it seems multiple exposures to be a little difficult.
Almost irrelevant. You're *in* the plane. The interior is a still subject and the horizon outside the window at cruising altitude ought to be fine for at least a quick 3 exposure burst.
Posted 2 months ago.
(
permalink
)
|
|
@Emteehed -- You wrote:
The thing about it is, do the HDR programs recover highlight details the same way? In my experience, no they don't. Photomatix wasn't able to produce the same image from just tone mapping it. So basically i have to break the RAW into several exposures and merge them back into an HDR file.
Here is something to consider: It's more likely an issue with the way Photomatix converts the RAW file, versus how it handles 8-bit images.
Unfortunately, they keep changing their conversion process and they don't always get it right. I've found that if I shoot a set of multiple exposure RAW files, I often get better results if I convert those into TIFF files and process them, rather than if I let Photomatix try to demosaic the RAW files itself. Theoretically, they should produce the same results. Hell, if you want to get technical about it, theoretically, processing the RAW's directly should give better results.
But it often doesn't.
This tells me that Photomatix has a flaw in its RAW processing, which probably has a lot more to do with people getting better results by faking multiple exposures, versus importing the RAW directly.
Originally posted 2 months ago.
(
permalink
)
Zuiun edited this topic 2 months ago.
|
|
Seems to me that some people still believe in some magic around the RAW files.
"The point of what I was saying was against people saying that you can't add what isn't in the single RAW. And to a certain extent, you can." Now please don't tell me that your camera sensor didn't already write the dirt details from the window into your RAW file!
So, all the details were already in the RAW file from the beginning on.
As Zuiun already stated there is no way that any software can get real details out of a RAW file that weren't captured by the sensor in the first place - period-.
Only magic or your fantasy can do that. Interpolations are no real details. Real increased dynamic range can only be added by merging multiple real bracketed shots.
Some software can reveal some details from the RAW files other software can't; it's as simple as that.
Using all kinds of cumbersome "HDR treatments" still will never produce a HDR image out of a LDR (low dynamic range) RAW file. Who still believes otherwise can just continue dreaming.
Posted 2 months ago.
(
permalink
)
|
|
i. love. these. threads...
Posted 2 months ago.
(
permalink
)
|
|
Wow. You are dense. If a channel is clipped, then its clipped. Now if i have software that can bring this channel back, interpolated or not, why wouldn't I take advantage of it?
The point is, it's still adding info that isn't in the raw file. It may not be 1-2 stops worth of light, but its still adding info not found in the file.
Now, no where in my search for an HDR software did I come across the companies talking about a highlight recovery system. So why would I expect the software to produce those details? I can't. And so something has to be done.
You may ask, why not just use the highlight recovery and add a lot of fill light when processing the RAW. Because the image will come out with a heck of alot more noise in the shadows. It won't look as natural and it will look like the Shadows/Highlights function in PS.
Just because it's not from real multiple exposures, doesn't mean that you can't fake the exposures with the RAW. To me, the only thing different is that the level of detail won't be nearly as clean or as good, but it's still making an HDR file from one RAW file.
This is hugely dependent on the scene you're photographing as well, as this shit will not work on a scene wit the sun in it or direct sources of light like street lamps.
Originally posted 2 months ago.
(
permalink
)
Emteehed edited this topic 2 months ago.
|
|
@Emteehed
"The point is, it's still adding info that isn't in the raw file."
"still making an HDR file from one RAW file"
Sorry, but if you still believe in magic, then maybe you just do not want to understand reality. If the sensor of your camera is not capturing certain details and writing them into a single RAW file, then where should the added information come from? Your treatment and software might interpolate some transitions between existing shades but it has no magic intelligence that can guess the correct details that were not captured by the sensor. And it is definitely not adding any dynamic range.
A couple of days ago your sensor captured the image in the airplane and wrote the data into a single RAW file. Now you claim that the information you reveal with your treatment and the software, like for example the dirt details, have not been originally written into the RAW file. That's the same as somehow believing that your magic software is bending the time-space continuum, looking back at the airplane window again and writing these data into your image and producing a HDR image from a single LDR RAW file. - dream on...
Posted 2 months ago.
(
permalink
)
|
|
@Emteehed -- you wrote:
Wow. You are dense. If a channel is clipped, then its clipped. Now if i have software that can bring this channel back, interpolated or not, why wouldn't I take advantage of it?
Name calling now. Priceless. So, what? Bad week at the office? Is that why you're so itching for a fight here?
The point that you are missing (or just ignoring) is that the data for the dirt smudges on your plane window *is* there to some extent. Your software didn't just say "hey, I think it might look neeto to invent some dirt here."
As I pointed out above, without your RAW file, it is impossible to know how recoverable those details were through other RAW processing methods. But the thing is, regardless of channel clipping, those smudge details -- to some extent -- *are* in the RAW file, even if they are not apparent in the initial exposure setting. Clipped is clipped and even DxO has to have *something* to work with from at least two of the three channels in order to interpolate its hightlight recovery.
You are not -- I repeat, NOT -- bringing information "back" into your clipped channel through the DxO method. There is no way any software can magically know what that information was in the scene in order to "bring it back." Either the data is there, or it is not. The interpolation process does add NEW data into the file, but NOT data from the actual scene.
No one is saying that you can't use software to interpolate data -- but that still is not *information from the scene.* it is a best guess about that information. It may look fantastic, but it is still not actual data from the scene. Period. End of story.
Bottom line: Can such software improve the highlights in an exposure? Sure. Are they accurate to the scene. *Not necessarily.*
That is the KEY difference between capturing multiple exposures for HDR and faking it through some other means. Multiple exposures will give you MORE ACCURATE details from the scene across a WIDER dynamic range than is possible from a single capture.
You wrote:
Now, no where in my search for an HDR software did I come across the companies talking about a highlight recovery system.
That's because highlight detail is already a byproduct of merging multiple exposures and tone mapping them.
You wrote:
You may ask, why not just use the highlight recovery and add a lot of fill light when processing the RAW. Because the image will come out with a heck of alot more noise in the shadows.
You mean like your airplane image?
You wrote:
Just because it's not from real multiple exposures, doesn't mean that you can't fake the exposures with the RAW.
No one is saying you *can't* do that. There is nothing wrong with using a single image and either processing it, or faking exposures from it and processing those. But that isn't HDR, it isn't an increased dynamic range, and the accuracy of the detail isn't the same as true multiple exposures.
You wrote:
To me, the only thing different is that the level of detail won't be nearly as clean or as good, but it's still making an HDR file from one RAW file.
No, it's making a tone map of the RAW file that has the HDR look. You can call it HDR if that makes you feel warm and fuzzy, but technically, it isn't.
Posted 2 months ago.
(
permalink
)
|
 |
People don't develop effects to create realism. Period. Tone mapping is just the newest in the chain of effects that people scream "ABUSE!" over time and time again. It just so happens that the HDR image packages out there use Tone Mapping as their most useful effect... and people like the results.... so they use the tool in as many ways as possible.
BTW, I'd guess that most photographers using HDR to create realistic images don't go broadcasting that they're using HDR. It would kinda detract from the magic of their art.
Realism in photography is simply an illusion anyway.
Posted 2 months ago.
(
permalink
)
|
|
NO!! I'm saying that the data is PARTIALLY not there. NOT saying that it's completely not there. You're assuming that I'm saying that the data isn't there at all.
This is why I said 'to some extent'.
And breaking the RAW file into separate images and merging them into an HDR file is what I would call making an HDR. PERIOD. It may be an HDR of a LDR scene, but it's still an HDR file. Because of the highlight recovery performed while separating the RAW it DOES and WILL come out with a better tone mapped image then just bringing that same RAW file into a program and tone mapping it.
Posted 2 months ago.
(
permalink
)
|
 |
I agree with you bYtE-sMaShEr, I am out to capture the best landscape images I can and if I am not tonemapping, I am HDRing it. If I explain what I did to non photographer they say I am cheating so I choose not to reveal the truth and just let them appretiate the images as one single capture.
Posted 2 months ago.
(
permalink
)
|
|
Yes! Please! Continue!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Posted 2 months ago.
(
permalink
)
|
 |
Bah... do what you want to do... criticism or not - no one is wrong. It's all about personal tastes.
HDR means High Dynamic Range... en.wikipedia.org/wiki/High_dynamic_range_imaging
In reality, it was invented at a time where many films didn't have the sensitivities of modern digital cameras. One of the simplest ways to do HDR imaging without stacking is using an S-curve in photoshop. it's a technique that I've been doing with many 48 bit scanned images in print for over 20 years.
It of course is more complicated that that. HDR is by all means an exaggeration of range to be represented on a substandard medium (namely your computer screen). While it is fair to criticize that it's over done, technically speaking what people are doing is exactly in the spirit of the original technique (which was used to shoot nuclear explosions to show both highlights and shadows which could never be accomplished with just one exposure).
However there are limitations to a single file giving you the right amount of information. Even if the dynamic range of the sensor can capture a huge range, over exposure can not be compensated. If your scene is poorly lit, you'll never get the highlights you need without going into the IR range which most digital cameras filter. Conversely, you can't recover the shadows if a scene is hot. Your sensors are overloaded with information that no adjustment curve can compensate for it.
So this opened the door for stacking or multi-exposure HDRs and tonemapping. By fixing several gamma points and then recombining images into a single image through various alpha channel mixing techniques, the HDR is faked to represent a range that is beyond reality. Based on this, technically speaking what you think HDR is today is actually psuedo itself.
Regardless of overdone or not, I prefer reality over the illustrated looks of HDR. But I appreciate it when it has been done tastefully.
My technical interest in image stacking isn't only in getting higher dynamic range, but to reduce noise for applications like night photography and astrophotograph. Whatever your reasons for HDR, I find it funny how people argue over it. Whether it's single HDR or multiple exposure HDR, it really doesn't matter as long as it accomplishes what the photographer intended. Love it hate it, it's hear to stay and no complaining will make people stop making it.
Posted 2 months ago.
(
permalink
)
|
|
@Emteehed -- you wrote:
NOT saying that it's completely not there. You're assuming that I'm saying that the data isn't there at all.
Here, I'll help you out with some quotes of yours to refresh your memory (I've emphasised the fun bits):
Copied directly from your initial post in this thread: Granted it's not going to be like using multiple exposures, but to me that's adding information that isn't in the initial exposure.
Another quote of yours: The point of what I was saying was against people saying that you can't add what isn't in the single RAW. And to a certain extent, you can.
And another: If a channel is clipped, then its clipped. Now if i have software that can bring this channel back,
One more: The point is, it's still adding info that isn't in the raw file.
You have stated multiple times that using DxO software allows you to add information into the file that isn't there -- which is 100% accurate according to the DxO specs. It's very clear that the software takes the existing data and interpolates NEW data that it thinks is appropriate.
But the two key things here are:
1) No software in the world can add real dynamic range information to a single RAW file that wasn't originally captured.
2) Any sort of highlight recovery that uses interpolation may produce good looking results, but isn't necessarily producing accurate results ("accurate" referring to the information existing in the scene). The new information added is, at best, a mathematical guess.
You wrote:
And breaking the RAW file into separate images and merging them into an HDR file is what I would call making an HDR. PERIOD. It may be an HDR of a LDR scene, but it's still an HDR file.
You can call it whatever you like. Call it a Fuzzy Bunny Slipper file if you want. What you call it doesn't alter what it is -- or in this case, isn't.
There is HDR -- which stands for one thing and one thing only: High Dynamic Range.
Then there is HDRi -- which is the 32-bit floating point file format.
What you are creating is an HDRi file, but the fact that it has been converted to a 32-bit file format doesn't give it a single bit more dynamic range information than what was present in that original RAW file -- no matter how many "exposures" you make from that single RAW.
"HDR of a LDR scene" is completely nonsensical. If you start with a low dynamic range file, you will end with a low dynamic range representation. It may be a tone mapped LDR representation, giving it some of that HDR "look," but it will still lack the detail of a tone mapped multiple exposure image.
To take that a step further, even if you take genuine multiple exposures of a low dynamic range scene, you will not magically gain a high dynamic range representation -- even after tone mapping, it will still be merely a fancy representation of that LDR scene.
Posted 2 months ago.
(
permalink
)
|
|
Why don't you quote me on this also??
"The point is, it's still adding info that isn't in the raw file. It may not be 1-2 stops worth of light, but its still adding info not found in the file."
From my experience, knowingly clipping highlights in one channel, highly dependent on the scene as well, most of the time I will be able to get from 1/3-1/2 stop of light recovered. Rarely 1 stop.
Originally posted 2 months ago.
(
permalink
)
Emteehed edited this topic 2 months ago.
|
|
@Zuiun
I guess it's a bit pointless to argue in length against ignorance. You are better with words and my english is not so good but technically we are expressing the same thing. We just don't believe in magic, we believe in facts, knowledge and logic.
@Emteehed
Last time....
Recovering "rarely 1 stop" doesn't make a HDR image out of a LDR RAW photo - even if you pack it into a HDRi format file. ....and again...if you still believe in your magic - just dream on...
Posted 2 months ago.
(
permalink
)
|
 |
how about everyone stops arguing and trying to convince others what HDR is for and meant to do? Its like your trying to say the paint brush was meant to only paint walls and not striking imagery. Some people like natural HDR, some like tonemapping..just let people do what they want. Its a tool and a technique, what we do with it and why is not in your control. wayne has options to visit other groups and what not which is what he chose to do. And keep shooting! :)
Posted 2 months ago.
(
permalink
)
|
|
@ Bret Konsdorf -- Hey, trying to tell people not to argue makes about as much sense as trying to tell people what HDR is and isn't for. ; )
For the record, I'm not trying to tell people what to *use* the HDR process for. Or even what they personally can call it. I'm just trying to explain what it is -- and isn't. It's all about knowledge.
And as that great American hero G. I. Joe says, "knowing is half the battle."
Posted 2 months ago.
(
permalink
)
|
|
I see what your saying Zuiun. There is a lot of confusion as to what HDR actually is. I just needed to vent a little :) not all directed at you haha sorry
Posted 2 months ago.
(
permalink
)
|
 |
I'm more tired of people whining about what is a 'true' HDR image, and other people complaining about 'overdone' HDR images.. it's a way for people to express their creativity in their images.
Stop your bellyaching, everyone.
Posted 2 months ago.
(
permalink
)
|
|
cal_gecko,
I understand how you feel. But, I'm getting tired of people excusing the misuse of a technique by calling it 'persoal expression'. It's laziness and being uwilling to actually learn about what they're attempting to do.
Posted 6 weeks ago.
(
permalink
)
|
 |
cant we all just just get along
Posted 6 weeks ago.
(
permalink
)
|
|
eminer...how do you know if its laziness?? maybe those that you speak of actually like the look?? Just like some like the look of thick brushes verses fine pens or pencil sketches.... I dont see how you can make that statement about art..it really confuses me.
Posted 6 weeks ago.
(
permalink
)
|
|
I'm with eminer and Zuiun
We are not discussing about the art here per se. What we are discussing is the technique behind the art. And techniquewise there is a certain set of logic, knowledge, hard facts, common procedures, science, laws of physics and conventions involved. No doubt, one might produce good art even if he does not understand what he is really doing techniquewise but this is not the point here. Confusing the precise technical definitions or not knowing them, maybe not understanding them doesn't prevent one from being an artist. However, claiming that it's not important to be precise and correct in terms of technique because it's just art anyways - or claiming you are allowed to mix all technical terms up, confuse them and define or re-define them however one likes "because it's art" might even be rather ignorant. And I'm with eminer1254 if he calls those lazy who might think that because it's just art anyways, they don't need to acquire deeper technical knowledge and learn the details and facts behind it and can then mix it all up,
Poste | |
HDR /
Discuss
