32 bit photoshop CS2?

Paul Furman wrote:

Very interesting. I did not know that. Is it only the HDR ‘plugin’ that works in 32 bit? I’m surprised this didn’t come up in that discussion I mentioned.

Here’s how I understand it:

CS2 has built-in support for 32 bit images. You do not require a "plug-in", although there are companies that make them. A 32 bit HDR image contains luminance levels that far exceed the luminance data that can be stored in 8 or 16 bits-per-channel image files. CS2 lets you make exposure and contrast corrections so that converting a 32 bits-per-channel HDR image to 8 or 16 bits per channel results in an image with the dynamic range (tonal range) you want. (Tonal mapping.)

CS2 also has an item under "File–>Automate–>Merge to HDR" that takes multiple 8 or 16-bit images and merges them into a single 32 bit HDR image.

I’d guess that no one raised the issue of 32 bit images in your other discussion because no cameras (as far as I know) yet support 32 bit images. My take on this whole question of 8-bits versus more is that most current display hardware can’t handle more than 8 bits. I guess the argument some people make is that you don’t need more bits than what you can ultimately display. That’s not true in my opinion. The extra bits provide lots more opportunity to manipulate and edit the image without losing colour resolution in the final 8-bit image. For example, there is no way that anyone could photograph two fully lit lightbulbs as a single 8-bit image and still maintain rich detail in both the highlights and shadows. That requires more bits to record the full luminance range and eventually for manipulation and tone mapping. This is despite the fact that the final result is still displayed as 8 bits.

Mardon wrote:

Paul Furman wrote:

Very interesting. I did not know that. Is it only the HDR ‘plugin’ that works in 32 bit? I’m surprised this didn’t come up in that discussion I mentioned.

Here’s how I understand it:

Thanks. Interesting. Roger Clark claims photoshop does sloppy math and does not get the best out of images but I guess that’s more an issue with the RAW converter, I’m not clear on that. Maybe CS2 does better in this regard. Can Camera Raw output 32 bit? It seems to me, not much value in 32 bit HDR if you can’t even percieve it with your eyes in a print. I often have the desire to capture high dynamic range images, this is rather important to my way of shooting but I think once you get all the data stretched they way you like, 8 bits should be plenty to display it.

CS2 has built-in support for 32 bit images. You do not require a "plug-in", although there are companies that make them. A 32 bit HDR image contains luminance levels that far exceed the luminance data that can be stored in 8 or 16 bits-per-channel image files. CS2 lets you make exposure and contrast corrections so that converting a 32 bits-per-channel HDR image to 8 or 16 bits per channel results in an image with the dynamic range (tonal range) you want. (Tonal mapping.)

CS2 also has an item under "File–>Automate–>Merge to HDR" that takes multiple 8 or 16-bit images and merges them into a single 32 bit HDR image.

I’d guess that no one raised the issue of 32 bit images in your other discussion because no cameras (as far as I know) yet support 32 bit images. My take on this whole question of 8-bits versus more is that most current display hardware can’t handle more than 8 bits. I guess the argument some people make is that you don’t need more bits than what you can ultimately display. That’s not true in my opinion. The extra bits provide lots more opportunity to manipulate and edit the image without losing colour resolution in the final 8-bit image. For example, there is no way that anyone could photograph two fully lit lightbulbs as a single 8-bit image and still maintain rich detail in both the highlights and shadows. That requires more bits to record the full luminance range and eventually for manipulation and tone mapping. This is despite the fact that the final result is still displayed as 8 bits.

Paul Furman wrote

Can Camera Raw output 32 bit?

It will save as 8 or 16, as I understand it.

It seems to me, not much
value in 32 bit HDR if you can’t even percieve it with your eyes in a print.

Actually, that’s not the main point of it.

I often have the desire to capture high dynamic range images, this is rather important to my way of shooting but I think once you get all the data stretched they way you like, 8 bits should be plenty to display it.

That *is* the point. You take multiple exposures of the same seen (2 or perhaps three is plenty). The idea is to capture the darker detail (by exposing for it) in one shot, and the highlights (by exposing for them) in another, and then the "merge to HDR" tries to give you the best of both, in a way the camera can’t capture on its own. It’s sort of like a much better split ND filter that doesn’t work along straight lines only.

http://www.luminous-landscape.com/tutorials/hdr.shtml

http://www.cambridgeincolour.com/tutorials/high-dynamic- range.htm

—
Randy Howard (2reply remove FOOBAR)
"The power of accurate observation is called cynicism by those who have not got it." – George Bernard Shaw

Randy Howard wrote:

Paul Furman wrote

Can Camera Raw output 32 bit?

It will save as 8 or 16, as I understand it.

It seems to me, not much
value in 32 bit HDR if you can’t even percieve it with your eyes in a print.

Actually, that’s not the main point of it.

I often have the desire to capture high dynamic range images, this is rather important to my way of shooting but I think once you get all the data stretched they way you like, 8 bits should be plenty to display it.

That *is* the point. You take multiple exposures of the same seen (2 or perhaps three is plenty). The idea is to capture the darker detail (by exposing for it) in one shot, and the highlights (by exposing for them) in another, and then the "merge to HDR" tries to give you the best of both, in a way the camera can’t capture on its own. It’s sort of like a much better split ND filter that doesn’t work along straight lines only.

I heard you cannot use multiple RAW conversions of the same shot though. That’s a shame because it means you are limited to shooting with a tripod. I can get a lot out of a raw file with different settings in two passes but it’s hard work merging them.

http://www.luminous-landscape.com/tutorials/hdr.shtml

< http://www.cambridgeincolour.com/tutorials/high-dynamic-rang e.htm>

Paul Furman wrote:

Can Camera Raw output 32 bit?

No.

It seems to me, not much value in 32 bit HDR if you can’t even percieve it with your eyes in a print.

It’s an intermediate format, not used for output.

I often have the desire to capture high dynamic range images, this is rather important to my way of shooting but I think once you get all the data stretched they way you like, 8 bits should be plenty to display it.

It is. The point is that you compress the range into something usable.

—
Jeremy |

Paul Furman wrote:

I heard you cannot use multiple RAW conversions of the same shot though.

There would be no benefit in doing so; the whole point of HDR is to get more range than a single shot is capable of capturing. If you only have one shot, you can just use the normal image.

I can get a lot out of a raw file with different settings in two passes but it’s hard work merging them.

Learn "curves" instead, and you’ll never have to do that again.

—
Jeremy |

Paul Furman wrote
(in article <Gxu5g.66260$>):

Randy Howard wrote:

Paul Furman wrote

Can Camera Raw output 32 bit?

It will save as 8 or 16, as I understand it.

It seems to me, not much
value in 32 bit HDR if you can’t even percieve it with your eyes in a print.

Actually, that’s not the main point of it.

I often have the desire to capture high dynamic range images, this is rather important to my way of shooting but I think once you get all the data stretched they way you like, 8 bits should be plenty to display it.

That *is* the point. You take multiple exposures of the same seen (2 or perhaps three is plenty). The idea is to capture the darker detail (by exposing for it) in one shot, and the highlights (by exposing for them) in another, and then the "merge to HDR" tries to give you the best of both, in a way the camera can’t capture on its own. It’s sort of like a much better split ND filter that doesn’t work along straight lines only.

I heard you cannot use multiple RAW conversions of the same shot though. That’s a shame because it means you are limited to shooting with a tripod. I can get a lot out of a raw file with different settings in two passes but it’s hard work merging them.

Well, the HDR merge just sort of cancels out the interpretations you took from a single raw image, if I understand it correctly. Yes, a tripod is the basic idea.

Apart from sports/action photography, just about anything worth an HDR process is worth carrying a tripod for in my opinion.

http://www.luminous-landscape.com/tutorials/hdr.shtml

< http://www.cambridgeincolour.com/tutorials/high-dynamic-rang e.htm>

—
Randy Howard (2reply remove FOOBAR)
"The power of accurate observation is called cynicism by those who have not got it." – George Bernard Shaw

Paul Furman wrote:

I heard you cannot use multiple RAW conversions of the same shot though.

that’s right; there is no point in using single RAW for HDR, as curves do better job when applied on single (RAW) image.

That’s a shame because it means you are limited to shooting with a tripod.

tripod is not strictly required. HDR may counterbalance small camera movements between shots (i.e. bracket exposition from hand is usually OK). Subject movement (between shots) is much worse problem – e.g. HDR will not produce good results from nature shots in a windy day.

B.

Randy Howard wrote:

It seems to me, not much
value in 32 bit HDR if you can’t even percieve it with your eyes in a print.

Actually, that’s not the main point of it.

theoretically it’s possible to capture multiple RAW images (ISO bracketing) from single exposure on CMOS sensor. Sadly, camera vendors that use CMOS are not innovative enough to even try it.

B.

Bronek Kozicki wrote
(in article <e37a2n$3f9$>):

Randy Howard wrote:

It seems to me, not much
value in 32 bit HDR if you can’t even percieve it with your eyes in a print.

Actually, that’s not the main point of it.

theoretically it’s possible to capture multiple RAW images (ISO bracketing) from single exposure on CMOS sensor. Sadly, camera vendors that use CMOS are not innovative enough to even try it.

Hmmm. "InstaBracketing (TM)" sounds like a plan. It would be nice to get the effect of HDR (with possible extended write times to store the results) in-camera. If it does happen, I’m guessing Nikon will NOT be the early adopter. 🙁

—
Randy Howard (2reply remove FOOBAR)
"The power of accurate observation is called cynicism by those who have not got it." – George Bernard Shaw

Hi,

I guess

the argument some people make is that you don’t need more bits than what you can ultimately display. That’s not true in my opinion. The extra bits provide lots more opportunity to manipulate and edit the image without losing colour resolution in the final 8-bit image.<<

I’m using a Canon 350XT shooting RAW. I open these images in PS CS2 as 16-bit images and tweak them before finally converting to 8-bit images for saving/printing. I think (belief system) that working on them in 16-bit mode offers a wider color gamut and superior output. But that’s only my "belief" system. I may be wrong.

Best,

Conrad
Camp Sherman, Oregon

There are some oddities in the images. Like the glow in the reflector image
or the blue sky tone gradations in the Photomatix example image on their home page.
But the images look good.

Jeremy Nixon wrote:

Paul Furman wrote:

I heard you cannot use multiple RAW conversions of the same shot though.

There would be no benefit in doing so; the whole point of HDR is to get more range than a single shot is capable of capturing. If you only have one shot, you can just use the normal image.

I can get a lot out of a raw file with different settings in two passes but it’s hard work merging them.

Learn "curves" instead, and you’ll never have to do that again.

Oh, I know curves but don’t have CS2 yet so can’t apply to raw. The Camera Raw adjustments do more than that though. With CS1, I can reveal wildly different parts of the exposure with different settings.

My point in this thread (and the previous 16 bit one) is maybe exporting to larger gamut or bit depth would allow better control in PS but it doesn’t look like it.

Paul Furman wrote:

Oh, I know curves but don’t have CS2 yet so can’t apply to raw. The Camera Raw adjustments do more than that though. With CS1, I can reveal wildly different parts of the exposure with different settings.

Yeah, CS2’s Camera Raw curves make that obsolete.

—
Jeremy |

In message ,
Jeremy Nixon wrote:

Paul Furman wrote:

Oh, I know curves but don’t have CS2 yet so can’t apply to raw. The Camera Raw adjustments do more than that though. With CS1, I can reveal wildly different parts of the exposure with different settings.

Yeah, CS2’s Camera Raw curves make that obsolete.

Besides being coarse in control, it is incapable of addressing anything outside of the blackpoint and whitepoint determined by the exposure, contrast, and brightness settings. It really isn’t much different than working with curves after the conversion. IOW, it is *NOT* a RAW curves tool.
—

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy

<<> <>>< <>>< ><<> <>>< ><<> ><<> <>><

wrote:

Yeah, CS2’s Camera Raw curves make that obsolete.

Besides being coarse in control, it is incapable of addressing anything outside of the blackpoint and whitepoint determined by the exposure, contrast, and brightness settings. It really isn’t much different than working with curves after the conversion. IOW, it is *NOT* a RAW curves tool.

Sure it is; it’s just not the one you want.

—
Jeremy |

wrote:

In message ,
Jeremy Nixon wrote:

Paul Furman wrote:

Oh, I know curves but don’t have CS2 yet so can’t apply to raw. The Camera Raw adjustments do more than that though. With CS1, I can reveal wildly different parts of the exposure with different settings.

Yeah, CS2’s Camera Raw curves make that obsolete.

Besides being coarse in control, it is incapable of addressing anything outside of the blackpoint and whitepoint determined by the exposure, contrast, and brightness settings. It really isn’t much different than working with curves after the conversion. IOW, it is *NOT* a RAW curves tool.

It sounds good, whay isn’t it *real* RAW curves? There are other adjustments in Camera RAW though, which can be used for multiple conversions with masking.