| Author | Thread |
|
|
03/02/2006 12:42:13 AM · #1 |
Would you share your personal experiences with Photoshop CS2's HDR feature? I've read some failure stories on using it especially when the subject isn't static in the photos so PsCS2 did a poor job at it.
My questions here is CURRENTLY more directed towards HOW we should setup the camera to take the originals before HDR? Use same aperture and take 2 shots or change the shutter speed? what about ISO? Would the EV compensation works just as well for HDR purposes?
Success stories would also be welcomed. Thanks |
|
|
|
03/02/2006 12:52:02 AM · #2 |
I'm interested in some tips also. I just recently upgraded to CS2. Seems like it's a real chore to work with having to take up to 5-7 shots and not make any movements with the camera/tripod otherwise it would screw everything up.
Message edited by author 2006-03-02 00:53:16. |
|
|
|
03/02/2006 12:55:31 AM · #3 |
First and foremost, a good solid tripod is a must. After that a remote is nice since you won't have to touch the camera while making the exposure. Timers can work well also.
As for how the exposure is changed, that all depends on the subject and what you are wanting to accomplish with it. Vary the aperture and the DOF changes, vary the ISO and the signal to noise level changes. These two will likely only be changed if there is something specific you are wanting to accomplish. Changing the shutter speed has the advantage of not affecting this type of image much. That is, multiple exposures are best when the subject is stationary, and stationary subjects don't usually care what the shutter speed is.
EV compensation doesn't change the exposure, it changes the way the meter reads the scene. That may sound like picking nits, but it is important to understand that aperture, shutter speed and ISO are the only changes that can be made to alter the exposure. EV compensation changes the way the meter reads the scene, and allows the convienence of thinking in terms of +/- stops of exposure -- but the changes are made to the three variables above.
Having said that -- Yes, EV compensation is a good way to adjust the exposures, just be aware of which of the three variables are being changed.
David
|
|
|
|
03/02/2006 12:57:21 AM · #4 |
Can someone perhaps post a HDR image that they made?
I read an entire article about HDR and "How to" in December's Digital Photography magazine, and am only exposed to the example pic (the church interior) they used.
Would love to see other attempts.
Message edited by author 2006-03-02 00:57:54. |
|
|
|
03/02/2006 01:19:37 AM · #5 |
David,
Somtimes HDR would also say it does not have enough data to perform the task. I've read somewhere that if someone used the same RAW file, converted it into 2 TIFF files, it dont always work too. Have you ever tried this approach? I was hoping this RAW-TIFF convert technique could get away with shakes or moving subjects? |
|
|
|
03/02/2006 01:27:29 AM · #6 |
I don't have CS2 - just CS, so I haven't played with HDR. I routinely bring shadows up and highlights down to improve the tonality. Most of what I have done involves making layers and adjusting the tones differently before blending them together again.
I recently got my D70, which opens the world of RAW processing to me -- but I haven't had a chance to play with this particular aspect of it yet.
I could mention my understanding of how it works, and the legality for challenges, but this thread covers those better than I would. It use to have an example image (a car if I remember correctly), but it has since been removed.
I will mention, in response to it not working at times, the success depends on the RAW file having the detail in the first place. If the RAW file doesn't have it, or doesn't have it without introducing a lot of noise, the is not much that can be done.
David
Message edited by author 2006-03-02 01:30:39.
|
|
|
|
03/02/2006 01:55:02 AM · #7 |
I've just finished reading that thread. Thanks for the link, David.
May I quote kirbic's words copied from that thread below:
"06/06/2005 09:00:42 PM
Here's what's legal for challenges:
You must start with one physical exposure, so bracketing and using multiple source exposures is out. You can, however take one RAW file, process it twice at different exposure settings, and combine those using layers, or "merge to HDR" in CS2, and use the result in Advanced Editing challenges only. Not legal for Basic Editing challenges."
So this means HDR is legal under Advanced Editing? Since kirbic is SC, so that's final, right? I might wanna give HDR a try (trying to learn only) on upcoming challenges.
Thanks all for the tips and guidance.
Message edited by author 2006-03-02 01:57:12. |
|
|
|
03/02/2006 04:20:11 AM · #8 |
Originally posted by crayon:
I've just finished reading that thread. Thanks for the link, David.
May I quote kirbic's words copied from that thread below:
"06/06/2005 09:00:42 PM
Here's what's legal for challenges:
You must start with one physical exposure, so bracketing and using multiple source exposures is out. You can, however take one RAW file, process it twice at different exposure settings, and combine those using layers, or "merge to HDR" in CS2, and use the result in Advanced Editing challenges only. Not legal for Basic Editing challenges."
So this means HDR is legal under Advanced Editing? Since kirbic is SC, so that's final, right? I might wanna give HDR a try (trying to learn only) on upcoming challenges.
Thanks all for the tips and guidance. |
Well, as the SC are quick to point out, unless it is in the Administrative Announcements forum they are most likely just speaking for themselves. However, that particular RAW editing technique was brought before the SC and ruled as legal -- and announced in this thread (scroll down, it's not the first question). So in this, it is definitely allowed, provided only one RAW source is used.
I expect this and other such items to be made more clear (and easier to find) in the currently underway rules rewrite.
David
|
|
|
|
03/02/2006 05:16:21 AM · #9 |
| Sorry for my ignorance but what is HDR? |
|
|
|
03/02/2006 06:38:56 AM · #10 |
High Dynamic Range. Dynamic range is the difference between shadows and highlights in your photo. Our eyes can see a certain amount of difference (roughly 9 stops?)and cameras can see less (dependent on the camera). What the HDR function does in CS2 is I presume try to (by using multiple exposures? not sure don't have CS2) is increase the contrast range in the photo. This is so you can "fit" more exposure into the scene, from highlights down to the shadows in one shot.
Message edited by author 2006-03-02 06:40:11. |
|
|
|
03/02/2006 08:47:05 AM · #11 |
HDR - Hight Dynamic Range...
I don't use PS so cannot speak for that application. However, PhotoImpact 10 has a real nice HDR function that works like this:
You must use at least three (but more if possible) images taken at the same aperture but different shutter speeds so bracketing in Av mode is the simplest way to accomplish this. Because of this an HDR image made this way is not legal for challenges in basic or advanced. Anyway, an example of the process I follow is to mount the camera on a tripod, Av mode, focus and shoot images (either bracketed or manual shutter speed change for each exposure). In PI open the images and select HDR which will open the dialog box to edit and combine the images into one image - the HDR image.
Now while this is not legal for challenges, there is another way (in PI at least). If you create an HDR image from multiple images once, you can create and save a camera profile from that process. Once this is done, you can create an HDR image from a single exposure using this profile. This would be legal for challenges. However, I have found that to create a useful profile at least 5 to 7 images are needed. I found that the best profile was created using 9 images of the same exposure.
Some images don't look 'better' as HDR images and some do. The starting image scene is the key. A relatively flat image as far as contrast and tone range will not produce a good HDR image. An image that has a lot of natural contrast and tone range (between shadows, highlights, colors) will produce a more dramatic final HDR image. If done corrrectly though, most images will look 'improved' after the process. |
|
|
|
03/02/2006 08:59:15 AM · #12 |
| Sounds interesting, is there an equivalent for the gimp? |
|
|
|
03/02/2006 03:43:55 PM · #13 |
Originally posted by bluenova:
Sounds interesting, is there an equivalent for the gimp? |
HDR is an automation of the by-hand process of layering multiple exposures, masking if desired and blending them together.
David
|
|
|
|
03/02/2006 09:17:50 PM · #14 |
Originally posted by crayon:
David,
Somtimes HDR would also say it does not have enough data to perform the task. I've read somewhere that if someone used the same RAW file, converted it into 2 TIFF files, it dont always work too. Have you ever tried this approach? I was hoping this RAW-TIFF convert technique could get away with shakes or moving subjects? |
Tried it with Photo Impact, and it didn't work. But its better to just use Rawshooter Pro and play with the shadow/highlight contrast, fill ligght, curves, etc. I think that you can do almost as much with a single RAW file as with a set of three exposures, bracketed at -2, 0 +2.
|
|
|
|
03/02/2006 09:34:23 PM · #15 |
Originally posted by hankk:
I think that you can do almost as much with a single RAW file as with a set of three exposures, bracketed at -2, 0 +2. |
Is that the normal bracketting you use? Is it sufficient for most situations to capture both the dark/light details in a high DR scene? Thanks |
|
|
|
03/02/2006 09:58:39 PM · #16 |
Originally posted by hankk:
Tried it with Photo Impact, and it didn't work. |
That is because PhotoImpact requires that the aperture setting of the images be the same but the shutter speed to be different in order to work properly. That is why you cannot do it with multiple conversions from one RAW image.
|
|
|
|
03/03/2006 09:44:25 AM · #17 |
Originally posted by crayon:
Originally posted by hankk:
I think that you can do almost as much with a single RAW file as with a set of three exposures, bracketed at -2, 0 +2. |
Is that the normal bracketting you use? Is it sufficient for most situations to capture both the dark/light details in a high DR scene? Thanks |
That's as much as the 300D gives me. PI will do some registration, so if I use burst mode, I can handhold a 3 shot burst and it will be close enough for PI to align the shots. (assuming the shutter speed is fast enough to handhold.)
I'd like to try -3, 0, +3, or even five shots, each two f-stops apart, but that's a lot of work (and you have to use manual mode and calculate anything over 2 f-stops away from 0, because that's the limit of the 300D's meter). |
|
|
|
03/03/2006 09:47:52 AM · #18 |
Originally posted by Alienyst:
Originally posted by hankk:
Tried it with Photo Impact, and it didn't work. |
That is because PhotoImpact requires that the aperture setting of the images be the same but the shutter speed to be different in order to work properly. That is why you cannot do it with multiple conversions from one RAW image. |
Just did some reading on other issues, and learned more about layer masks and layer transparncy. That may be a good alternative when I absolutely positively need a great shot and can't get HDR to work, or when I absolutely positively feel the need for total control...
|
|
|
|
03/03/2006 01:21:30 PM · #19 |
Has anyone seen any good, not nasty, looking results from HDR that wouldn't be possible with one decent exposure ?
All of the results I've seen just look wrong. I think we need HDR displays before HDR images will ever look right, but I'd be happy to be proven wrong.
Could just be culturally imposed expectations of what a photo 'looks' like though. |
|
|
|
03/03/2006 05:36:22 PM · #20 |
Originally posted by Gordon:
Has anyone seen any good, not nasty, looking results from HDR that wouldn't be possible with one decent exposure ?
All of the results I've seen just look wrong. I think we need HDR displays before HDR images will ever look right, but I'd be happy to be proven wrong.
Could just be culturally imposed expectations of what a photo 'looks' like though. |
The Ulead tutorial at //www.ulead.com/learning/pi/pi10_01_2.htm gives some good (although somewhat contrived) examples. |
|
|
|
03/03/2006 07:28:06 PM · #21 |
I use a software called "photomatix" and it works great. Better than CS2 in my opinion. I shoot alot of indoor images. This application lets me take low exposure images so that you can see out the windows with lighter exposure images for the interior. Once the images are merged, it is a great effect. Nice inside exposure without blown out windows.
|
|
|
|
03/04/2006 07:37:32 PM · #22 |
Today we finally managed to find a sushi restaurant with a sushi TRAIN! Quite a rarity around Seattle. I was so thrilled that I took a photo of the place.
The lighting was awkward - lots of glass at the front made that part very bright, whilst the back was quite dark.
Thankfully I took the photo in RAW (as well as jpg).
Back home, I made two versions of the RAW file - one light and one dark. Then I tried to figure out how to best blend them.
PSCS2's HDR feature didn't work - it told me I didn't have enough dynamic range in them.
So I tried one via photomatix as per David's suggestion, and one manually via layer mask.
I know the photo is nothing amazing, but I'm stoked that I have learnt enough to be able to do this at all.
1) the original jpg ..... 2)blending via photomatix ....3)manually in PS
   |
|
|
|
03/04/2006 08:20:43 PM · #23 |
HDR displays are here: //www.brightsidetech.com/products/dr37p.php
Almost all HDR->LDR converted images look like ass because you end up reducing the contrast to get the appropriate DR compression. Dual exposures with single/weak split dynamic range shifts give somewhat pleasing results. Some useful research (without useful implementations to date) has been done in HDR compression:
//www.cs.huji.ac.il/~danix/hdr/results.html
Just to correct the previous statement:
"HDR is an automation of the by-hand process of layering multiple exposures, masking if desired and blending them together."
This is not really about HDR at all, but about the common technique of multiple exposures combined in a simplistic way to yield apparent greater dynamic range than was captured. There is considerable literature on the subject. There are even HDR cameras around, but they are very expensive.
It is not unusual to take LDR cameras and build HDR profiles by varying sample exposures of a constant HDR scene. Subsequent LDR sequences can then be used to create a true (or truer) HDR image. Unless you have an HDR viewer (or display), the HDR image is always converted to an LDR. In some cases, it's a simple LDR slice with no compression done at all. Some of the more complex conversions attempt to compress most, if not all, of the HDR image, but as mentioned earlier, the contrast reduction produces rather ugly results.
Basically, you can't look directly at the sun and the crack of your ass at the same time; HDR to LDR compression attempts to bridge that gap. |
|
|
|
03/04/2006 10:18:36 PM · #24 |
| There's an article on HDR in the current issue of Outdoor Photographer |
|
|
|
03/05/2006 08:50:06 PM · #25 |
Not sure what I did wrong, but I really need help with the PS CS2 HDR function. I blended 3 images together, each differed by +1EV and -1EV respectively, and the resulting HDR image is a blown-out whites, and very dark (like black?) shadows! Entirely opposite of what I'm trying to achieve! grrr...
Message edited by author 2006-03-05 20:50:34. |
|
