Lightroom 5 vs DPP 3.13.10 - is this typical?

OK, I've purchased LR5 in the believe that I can use it to get more out of my photos, and since I knew it was arriving sooner than later I downloaded the trial version from Adobe (after a bit of run-around on the website). Probably could have done that a while ago, but I was lazy.

Here is a photo I edited in LR5. It's a bit of a mundane subject, but I'm fairly happy with the result of an evening shot experimenting with high-ISO and ETTR (T3i, 17-55 @ 28mm, f/2.8, Av mode with +1/3 exp resulting in 1/15 sec, ISO 3200), exported as jpg:

I'd say that most of us who have compared are persuaded that the Lightroom Highlights "recovery" tool is noticeably more effective than the DPP recovery capabilities. I haven't done a thorough comparison of all the various tools, so I'll just stop there!

If you post the raw file, then DPP users may try to see what is possible to do with DPP.
BTW it seems you have applied more sharpening in the LR conversion.

DPP is not intended to be a post-processor, so no wonder you cannot do what LR can do. It has a different role for me as a means of getting CS2 to "see" my raw images.

I'd be seriously annoyed if LR could not do more than DPP.

I'd say it is entirely typical. It is the reason I would never go back to DPP (left it 2008 for LR2). LR4/5 puts on your plate every bit of image data recorded by the sensor while DPP does the opposite, in order to prevent false colors in soft-clipped highlights (one or two channels clipped, but not all three) it intentionally clips about 1/4 of a stop off the top. It is possible to get fuller DR in DPP by doing two conversions; a normal one and a linear one and then blending them through a luminosity mask so that you are using only the highlights from the linear conversion - but that is a lot of extra work to do what LR is doing automatically.

Lowner wrote in post #16206441
DPP is not intended to be a post-processor, so no wonder you cannot do what LR can do. It has a different role for me as a means of getting CS2 to "see" my raw images.

Fair enough, but what arrives in PSCS2 is a truncated version. Highlight detail that wasn't included in the Raw conversion is gone forever.

yb98 wrote in post #16206428
If you post the raw file, then DPP users may try to see what is possible to do with DPP.
BTW it seems you have applied more sharpening in the LR conversion.

(my emphasis)
Yes, I did apply some clarity (I feel like John Anderton in Minority Report when I say that), because overall sharpening didn't do what I wanted. And of course that is one of the things LR brings to the table that DPP doesn't have - I knew about that difference, so I wasn't surprised when I found it effective.

I think tzalman has described it best for me: LR5 does a better job at using the entire dynamic range of the raw pixel/sensel information. That was what I was trying to figure out with the blown-out highlights of the lights in the water fountain.

I guess I'm just astonished that DPP would purposely throw away dynamic range that LR so easily utilizes.

Eh, I guess I'm a convert then. LR FTW. (Used non-sarcastically. Does anyone really use "FTW" any more?)

shinksma

shinksma wrote in post #16206775
Eh, I guess I'm a convert then. LR FTW. (Used non-sarcastically. Does anyone really use "FTW" any more?)

shinksma

Here in POTN it's more common to use

Can you please post the RAW file someplace accessible?

To say that DPP is 'throwing away dynamic range' is probably not quite correct. LR is no doubt doing some manner of local contrast enhancement that makes it look punchier.

Geonerd wrote in post #16208782
Can you please post the RAW file someplace accessible?

To say that DPP is 'throwing away dynamic range' is probably not quite correct. LR is no doubt doing some manner of local contrast enhancement that makes it look punchier.

I'm afraid it is correct. Let me explain: There is a phenomenon known as "False Highlight Color". Imagine a highlight color in which the the proportions of primary colors is that red and green equal X and blue is 1.25X. If you expose the shot so that red and green are 248 (8 bit scale), then blue should be 310 (248 X 1.25). But 255 is the maximum number that can be rendered in 8 bits, so the color is written as 248/248/255. I.e., it is a false color and the blue channel, being clipped, has lost data. The easy solution to the problem is to cut the data at a point below the top, where the primaries are still in the proper relationship, one to another. That's what DPP does, chopping the sensor data at about 12,500 instead of the full 16,000 (14 bit), which is a loss of about 0.25 stop. The fancier solution is to scale the highlights down appropriately and write data from the unclipped channels into the clipped channel.

Geonerd wrote in post #16208782
Can you please post the RAW file someplace accessible?

I don't have a personal website for uploading non-image files, all I have is a photobucket site. I've never had the need before, and I'm probably not going to get the chance to figure out what I want out of a dropbox or similar site to set one up any time soon. I'll post here if I do get the CR2 file uploaded somewhere.

Geonerd wrote in post #16208782
To say that DPP is 'throwing away dynamic range' is probably not quite correct. LR is no doubt doing some manner of local contrast enhancement that makes it look punchier.

I was just going on what tzalman (Elie) had posted previously, since it seemed to agree with what I experienced: i couldn't do anything to extract out the relative local contrast for the blown-out-looking underwater foutain lights using DPP to get the same result as LR5. i tried dropping the exposure, slamming contrast all the way up and down including the shadow and highlight controls, using the RGB tab to do the same as well - nothing could recover that level of apparent detail that LR5 did automatically. It truly looks like DPP just truncates the data beyond a certain dynamic range.

I'm happy to be shown that it can be extracted using DPP, but I'm also quite convinced that LR5 makes it easier, so I'm going to stick with the tool that makes it easier.

shinksma

Have you tried the linear mode of DPP ? sometimes (not always) it can do miracles !

shinksma wrote in post #16210193
I don't have a personal website for uploading non-image files, all I have is a photobucket site. I've never had the need before, and I'm probably not going to get the chance to figure out what I want out of a dropbox or similar site to set one up any time soon. I'll post here if I do get the CR2 file uploaded somewhere.

I recommend the site YouSendIt.com. It's a hassle-free site for uploading single files -- you give your email address as the "Recipient" and upload your Raw file. They email a notification to you with a link, you can post the link here, and we can download your file!

yb98 wrote in post #16210221
Have you tried the linear mode of DPP ? sometimes (not always) it can do miracles !

I hadn't but I just tried it on a different photo (since I'm not at my home workstation...), and it darkens the overall image and removes the contrast adjustment tools. It looks like it would be good where the whole image is overexposed, not images with a few "hot" areas.

Unless you mean a different linear mode than the checkbox below the contrast sliders on the raw tab.

I'll see what that does when I get a chance on the image above, but I don't have much expectations based on what I just saw.

shinksma

YouSendIt is now HIGHTAIL:

https://www.hightail.c​om/