New user, having trouble with HDR merge

[kruuth]

This is my first post, and I searched a little but didn't really find any help with this one.  I have an image that I snapped awhile back, of the front of the church at Cordoba, Spain.  I'm trying to do an HDR merge by adjusting a few things in the image, in this case, brightness, contrast, levels, saturation, etc.  No mater what I do I get the same "there is not enough dynamic range..." error.  I'm not too sure what to do, and I don't have multiple images.  Can someone give me a hand.  Thanks.

[DarkPenguin]

The PS HDR merge thing?  I'm 90% sure that it requires separate exposures.  One exposure with different conversions will not work.

You might want to try it using layers and a layer mask.

Or just use the one image with shadow highlight.

[kruuth]

I see.  There was a tutorial here:

http://www.vanilladays.com/hdr-guide/#comment-8416

that said I could do it with one image.  I was hoping to do that, as there is some motion I would like ot preserve.

The PS HDR merge thing?  I'm 90% sure that it requires separate exposures.  One exposure with different conversions will not work.

You might want to try it using layers and a layer mask.

Or just use the one image with shadow highlight.
[a href=\"index.php?act=findpost&pid=109074\"][{POST_SNAPBACK}][/a]

[Ray]

I see.  There was a tutorial here:

http://www.vanilladays.com/hdr-guide/#comment-8416

that said I could do it with one image.  I was hoping to do that, as there is some motion I would like ot preserve.
[a href=\"index.php?act=findpost&pid=109078\"][{POST_SNAPBACK}][/a]

The author of that tutorial on Photomatix wrote he didn't like the result. You can blend different versions of a single RAW file in oder to get the maximum dynamic range that's contained in the file, but whatever you do, you cannot extract more dynamic range than is contained in the file.

It's really a matter of choosing the method that gives you the best result and the most control. One method is to do a conversion with EC at say +2; another conversion with EC at -2, then blend the two images as though they were separate exposures, along the lines of Michael's tutorial on digitally blending images.

On the other hand, depending on your skill in Photoshop, you will get at least equally good results from just one conversion in 16 bit mode and ProPhoto RGB color space. To do a conversion that retains the maximum dynamic range that was captured, you need to convert with shadows, brightness and contrast at zero as well as the appropriate EC adjustment to retain as much highlight detail as you can.

I'm not sure how much advantage there is in setting contrast at -50.

[jbrembat]

On the other hand, depending on your skill in Photoshop, you will get at least equally good results from just one conversion in 16 bit mode and ProPhoto RGB color space. To do a conversion that retains the maximum dynamic range that was captured, you need to convert with shadows, brightness and contrast at zero as well as the appropriate EC adjustment to retain as much highlight detail as you can.

Sorry, it is not HDR. HDR must provide a tonemapping to compress the huge dynamic range  for viewing (or printing) on LDR (low dynamic range device).

ProPhoto color space is able to preserve all the colors... and more (it goes beyond visible), but the tonemapping performed using the color profile is poor.

Tonemapping must compress the dynamic range trying to preserve the apperance of the original scene.

Only one shot in raw, developed at different exposures, may be enough or not depending on scene dynamic range. Camera sensor is not HDR. Fuji super CCD is the first sensor trying to increase the dynamic range.

Jacopo

[Ray]

Sorry, it is not HDR. [a href=\"index.php?act=findpost&pid=109095\"][{POST_SNAPBACK}][/a]

I never said it was. The point I was trying to make is that programs like Photomatix (or HDR if it would work with different conversions of the same RAW file) cannot provide greater dynamic range than has been captured in that single file. There is no ultimate advantage in making dual or triple conversions. It's just a different method of getting the desired, processed effect.

[jbrembat]

Hi Ray,
I agree: there is no way to increase the dynamic range captured.

There is no ultimate advantage in making dual or triple conversions. It's just a different method of getting the desired, processed effect.

I don't agree.
The point is that when you develop your raw data, if your image has a strong contrast you must choose where to lose details, in shadows or in highlights.
Developing for multiple exposures and using a good tonemapping sw you can get better results (more details in shadows and highlights too).

So, the best HDR is got using multiple exposure, but getting multiple exposure from a single shot may be good too.

Of course, in the latter case you cannot go beyond the captured dynamic range.

Have a look at   http://www.photoresampling.com/hdr_eng.php

Jacopo

[Ray]

I don't agree.
The point is that when you develop your raw data, if your image has a strong contrast you must choose where to lose details, in shadows or in highlights.
Developing for multiple exposures and using a good tonemapping sw you can get better results (more details in shadows and highlights too).

[a href=\"index.php?act=findpost&pid=109117\"][{POST_SNAPBACK}][/a]

Hi Jacopo,
I used to think like that, but I don't think it's necessarily true! 16 bits per channel is capable of 'holding' a higher dynamic range than the 12 bit RAW data from most DSLR cameras. If you get your adjustments right in ACR, you'll recover the full dynamic range of the image with a single conversion.

It's then simply (perhaps not so simply) a matter of which approach in PS you are going to use to lighten the shadows, preserve the highlights and create the contrast and saturation which you think best represents your vision of the scene.

[jbrembat]

It's then simply (perhaps not so simply)

I think you get the right answer....

[Ray]

I think you get the right answer....
[a href=\"index.php?act=findpost&pid=109127\"][{POST_SNAPBACK}][/a]

An easy technique is to Ctrl left click on RGB channels, which selects the highlights, then select the inverse (short cut - Ctrl+Shift then I); use 'layers adjustment', curves and/or channels, opacity say 80%. This method protects the highlights from any brightening of the shadows and midtones.

There's no adjustment layer for unsharp mask, but you can use the same procedure then go to unsharp mask, selecting a small 'amount' and a big 'radius'. This increases local contrast whilst preventing broad halos around edges, something one often gets when digitally blending images manually.

[jbrembat]

An easy technique is to Ctrl left click on RGB channels, which selects the highlights, then select the inverse (short cut - Ctrl+Shift then I); use 'layers adjustment', curves and/or channels, opacity say 80%. This method protects the highlights from any brightening of the shadows and midtones.

There's no adjustment layer for unsharp mask, but you can use the same procedure then go to unsharp mask, selecting a small 'amount' and a big 'radius'. This increases local contrast whilst preventing broad halos around edges, something one often gets when digitally blending images manually.
[a href=\"index.php?act=findpost&pid=109145\"][{POST_SNAPBACK}][/a]

I'm not an expert of photoshop. But I think you are writing some method to adjust a "developed" photo. Tone mapping at this point is gone. A simple gamma correction was applied. Are you trynig to enhance shadows/highligths in a "developed" photo?
If so, you can, but you lost the HDR train.

Jacopo

[Ray]

If so, you can, but you lost the HDR train.
[a href=\"index.php?act=findpost&pid=109162\"][{POST_SNAPBACK}][/a]

But the original poster wrote that the HDR train doesn't work with variations on a single image, in PS. If you want to expand the dynamic range of a single image, you simply can't, no matter how many conversion variations you make.

[jbrembat]

But the original poster wrote that the HDR train doesn't work with variations on a single image, in PS. If you want to expand the dynamic range of a single image, you simply can't, no matter how many conversion variations you make.
[a href=\"index.php?act=findpost&pid=109163\"][{POST_SNAPBACK}][/a]

This is a PS problem, not a HDR problem!
As I said, HDR is scene dependent, so if camera sensor was able to record all or most of the dynamic range, you can recover an image that looks similar to the scene your eyes saw at shot time. If PS is not able to recover....., I reapeat, is PS problem.

Jacopo

[DarkPenguin]

I thought the OP was referring to a specific feature of photoshop.