DPP vs. Bridge (ACR) Accuracy Discussion between Pros

Note: whenever I say "Bridge", I mean ACR...

Several times, I have started to put together images that show the differences between Bridge (ACR), DPP and CS2/PS. For this discussion, I have not bothered and thought it would be a good topic to discuss amongst seasoned pros. (I actually wanted to take the time to develop a mouse-over flash presentation that shows the difference directly)

First, let me state that I have been a graphic arts professional for over 20 years. I have been in the digital mainstream for over 12 of those years. So, I know a little about system color calibration and the correct ways to use programs, drivers, settings, etc. (note: I am not saying I know everything) I also have absolutely no problem getting DPP and PS perfectly in sync. (to what I saw originally) The images between the two are 100% identical. (the way they should be) I should also say that if I never see the DPP preview that Bridge and PS are in sync. (to me, but darker and subdued)

However, since day one of owning my 30D and DPP, I have absolutely loved the fast workflow for general base editing, overall vividness and accuracy of the 'what I saw' that I get from using Digital Photo Professional (DPP) over Bridge (ACR). DPP seems more true to life and vivid as I remember seeing the actual scene or subject. Obviously, as a graphics arts professional, I have enjoyed using PS for many years (since version 2 actually), and most recently, Bridge in the past year for workflow features. However, the accuracy between Bridge (ACR), DPP (or specifically, the CR2 file itself) and PS in regards to photographic files and their default color, brightness, accuracy, etc continue to baffle me.

Bridge (ACR) has a tendency to subdue images when it opens them. This is a fact. I have personally disabled all of Bridges "auto adjustments" in hopes of seeing the same image I would see in DPP on default camera shot settings. The problem is, it's impossible. (without faking it by setting your own "changing adjustments" in Bridge) It does get closer, but still is not as accurate in color, or mainly brightness, as the original DPP preview and what I remember.

I shoot in sRGB. I edit in sRGB. Why? I use SmugMug and this is the way they are set up for consumer printing. (among many other 'photo' printers online) By using their profiles, prints are accurate. If you do not have a working background in offset printing and can explain to yourself the difference between why you choose AdobeRGB1998 over sRGB then do not bother to respond to this thread. (your time will be wasted) This discussion will be based on sRGB editing only. (if you are currently using AdobeRGB1998 and do not do CMYK offset or commercial printing, you might be interested in reading SmugMug's description of why you should be in sRGB mode... it's very enlightening to those who don't have a background in printing and are mainly photographers... this goes for all you in-house Epson owners too )

The bottom line is Bridge should not look any different than DPP in a default image preview (again, default shot settings) when the color settings, space, etc are identical. So, why does it? Is this a software issue with Bridge and how it reads the original file? Or is this a problem of DPP being 'too' vivid? Personally, I think DPP is much closer to accuracy regarding the actual subject, color, lightness, etc than Bridge (ACR) is.

Professionals, run your own tests... don't worry about print output at this time... lets only discuss why we think the viewing accuracy is different in Bridge, regarding the original image at it's default camera shot settings in comparison to the DPP preview of the same thing. If you feel Bridge is more accurate in regards to the original subject, color, brightness, etc - feel free to discuss that as well.

I challenge you to not reply until you have tested this yourself. I am not looking for a debate here; but a discussion with possibly a conclusion on where we need a solution.

Discuss...

Well I have compared the same images in DPP, Bridge/ACR and Lightroom

For my 1DsMkII images I like DPPs initial rendition the best, although I usually need to add a touch of contrast and saturation for taste. Lightroom comes in second and ACR third.

And I agree ACR seems to mute the color down somewhat and I usually have to bump the image a touch more in PS to get a final that looks like what I got out of DPP in the first place.

Interestingly from playing with lightroom beta, it seems to do very good on some images - however as it is slower then dirt so far, it makes it fairly useless to me.

Meanwhile in a previous comparison between PS 7 w/ACR versus DPP with my 10D, ACR did a better job with those images as well as my D60 images then DPP did although the D60 images are basically tie.

It is of course quite possible to tweak the images to get them to match in output, but DPP seems to hit it right more often with less effort. Also by this point (2-years) I have enough experience to batch process in PS CS2 to get best results when I need it.

So in short my experience basically mirrors yours with my 1DsMkII, but not quite with my 10D/D60 and I suspect either the digic processor or Adobe got lazy between 10D and 1DsMkII.

Just my experience,

Longwatcher wrote in post #2158284
...I like DPPs initial rendition the best, although I usually need to add a touch of contrast and saturation for taste... I agree ACR seems to mute the color down somewhat and I usually have to bump the image a touch more in PS to get a final that looks like what I got out of DPP in the first place... quite possible to tweak the images to get them to match in output, but DPP seems to hit it right more often with less effort...

I agree with that 100%. (sorry for the editing)

I just went through this process two weeks ago.

I did a three day shoot for an upcoming Stunt Junkies show in very challenging lighting conditions (Money shot was one chance, 100mph subject in the rain, in a dark canyon at 250 yds). I had approx 2500 RAW’s from the event to sort through. After an hour of coming up with pathetic conversions using ACR I was ready to take a baseball bat to my camera and computer, literally!

I opened DPP to take a closer look at my EXIF. I wanted to study how I could possibly suck so bad as a cameraman and find out why my settings were so far out of whack. DPP shows the camera settings were exactly where they should have been and histogram confirmed. I’m shaking my head, “WHAT AM I DOING WRONG!”

I opened one of the photos in DPP, BAM! It’s PERFECT! I mean it’s absolutely DEAD ON PERFECT!!!!!!!!!!!!!!

Convert it to .tiff, PERFECT! (ok, just a little contrast +3-4) but I didn’t need to touch it after that. It was a night and day difference from ACR.

And all this time I was blaming myself.

Adobe has a lot of work to do. If their new Lightroom program is using the same process as ACR, I won’t even look at it.

A mouse over would be the perfect example. I would love to release these photos for that purpose but the show does not air until January. I would be in deep trouble if I posted them.

However, I would love to see other examples. You are on to something extremely important, step on up amonline.

P.S. An acknowledgment from Adobe stating, “Yeah, we know. We’re working on it” would be something so very special, especially since they are the ones who are trying to take over the world with DNG.

I identify with you Baz...

I have said all along that I believe Adobe truly does not have the correct info from Canon to handle the files correctly. I feel that Canon may be holding back the "magic mix" because they want to develop DPP to more of a high end consumer level application. This wouldn't surprize me a bit.

I may do the flash thing sometime soon... I have to find the right suject to do it with that truly shows what this thread is all about.

I'm new in acr process of raw images and notice the same problem. After read some threads in this forum looks like acr profiles needs to be tweaked a bit.

Here is the link for profiling acr http://people.csail.mi​t.edu/ericchan/dp/acr/​.

Seems like to much effort, but maybe you could give it a try.

Thanks dino... I'll definately check that out. It concerns me that such an application would have to be profiled so deeply when glancing at that info. I have seen mention of profiling in the past regarding this issue. It seems to be a solution that's been put in the mix to correct a problem really.

I look forward to reading that... thanks again!