I'm joining this discussion really late, but there are so many things that have been posted that are quite off besides the matching of your jpgs (or should I say "previews")... I wanted to add my two cents...
In regards to the OP's very first few posts; you're not going to achieve exactly what you get from the camera jpg because so many things are taking place in that algorithm. First, the camera is guessing at the best WB or one you've supplied. Then it will apply a slight auto-level adjustment along with any additional parameters you may have set. Finally, it's compressing all that info to an unknown quality level that will differ from the computer that you'll find nearly impossible to duplicate because you truly don't know what all took place. What we do know is that the quality is not equal to DPP's "10" given the splotchy jpg artifacts. My guess is something in the 6-8 range.
davidcrebelxt wrote in post #2469567
Zoombrowser (Canon's software) is the only one I can find that will produce images matching the in-camera jpeg by default. DPP (even though it is Canon's software too) does NOT.
On that note, I was wondering if anyone here has found what parameters to adjust in DPP to best match? I tend to shoot with Parameter 1, which boosts sharpness, saturation, and contrast... but I can't pin that down in DPP to match the in camera jpg.
PacAce wrote in post #2469568
The image coming out of DPP by default may not be 100% exactly like the JPEG image coming out of camera but it's so darn close that it'll be hard to tell them apart unless you have the two images sitting side by side to each other. Even the images coming out of IB or ZB are slightly off although I do admit it is closer to the in-camera JPEG than the one from DPP is.
If you do not match your chosen camera jpg algorithm, (which is nearly impossible as the camera plays with the levels upon compressing) then they are not going to even get close. If you DO match the algorithm and then export the file as a jpg (not just previewing or transfering it) then it will match nearly 100% outside of levels. (which will affect your whites) I'm guessing you are trying to achieve the jpgs color, sharpness and saturation, but not the lower jpg quality. I say it will match nearly because no matter what, the compression is going to differ slightly from device to device. It's fact that the compression at the computer will be 100 times better and produce a much higher quality image in the end. So, I understand your desire to use the jpg as a basis. In the long run you'll figure out how to set the camera parameters the way to only need the RAW.
Also, no one knows what level the in-camera jpg compresses to as it differs from the compression at the computer. This is where your splotchy color artifacts start showing up. Regardless, the Raw will always show blown out whites if you blew them out on the exposure. Just take a look at the jpg+Raw histograms in comparison... better yet, click the "auto-level" button and watch what happens. Look familiar?
davidcrebelxt wrote in post #2469571
On your shots, the one on the far right has the most "pop"
I think I have my CM set right, here a screenshot: (I also do the comparison between RAW and JPEG right in DPP, so I thought it should be same.:
If you personally print from DPP then you should use a printer profile. This is how you achieve the WYSIWYG printed result. However, you should never use "perceptual". Always use "Relative Colorimetric". (keeping black point if you have the choice) Relative Colorimetric maps the white point to the final medium. It also scales your colors more accurately based on the white point. It's best for inkjet printing and soft proofing to yield the best results. Locking the black point makes sure your bottom end doesn't become muddy or inaccurate.
If you send ALL your images to PS/whatever for further editing/printing, then you should NOT choose a printer profile and handle that in PS where you'll want your WYSIWYG. I use the printer profile because most shots that are in-house printed are fine processed and printed from DPP for my clients. However, I leave it set that way because even when I send images to PS for further editing or SFX, it doesn't really matter as PS will strip the profile anyway while I work from a new soft point and it still looks identical based on my sRGB gamut.
Alternatively, if you use an online printer such as EZ Prints (or any vendor where you upload the images for a client's proof and ordering) then you should be getting their profile loaded so you see what they will be printing. You can soft proof through PS/DPP with their profile as well, but most vendors will NOT want you to embed the profile as they generally like to work with sRGB files and let their machines do they converting. One reason for this is that the client sees what they will be receiving correctly, but saved for monitor proofing. Bottom line - use the printer's profile if you want to see the final correctly - otherwise, use sRGB for web related use... just don't embed the profiles.
PacAce wrote in post #2469572
This is a screen shot what mine looks like side-by-side in DPP. Other than a very slight hint of reddish cast on the image from the raw file (right), the colors don't seem that far off. At least not enough to make the raw image looked washed out.
[ HOSTED PHOTO DISPLAY FAILED: ATTACH id 134553 does not exist. ] Again, this is comparing a "preview" to an already compressed and saved jpg. If you are re-opening a saved jpg (which the name did not show) then you are simply saving at a higher quality than the camera's algorithm. We're talking about the camera compressing to an unknown quality level, but it can still be duplicated in DPP. If you really want to compare the two, batch the jpg out and open them side by side in PS. You will then see that they are quite closer than you think. If you play with the quality level, you'll hit it 99.9%. Here again, you won't match the camera's algorithm when it comes to level adjustment.
davidcrebelxt wrote in post #2469574
First image is Raw, with Default DPP settings, second is jpeg from camera. 100% crops
Your comparing a jpg to a RAW file that is not matched yet. Raw's do not carry the jpg-ed algorithm to the preview. You would have to match the settings and save it out at the matching compression level - THEN compare it.
davidcrebelxt wrote in post #2469646
Yeah, that's whats really throwing me off. Before I did those screen grabs I made sure I reverted to as shot settings. Those screen grabs were straight out of DPP. That was the RAW editing preview - I never saved to jpeg yet, so both should be displaying in my sRGB workspace.
Yes, but the jpg processing in camera will correct some exposure or level problems. DPP is showing you what you truly captured. The jpg is the processed and corrected file from the camera. The in-camera jpg save runs a slight level adjustment. (test it yourself)
I don't think you guys are comparing the correct things with the correct things. All I've seen here are on-screen previews of RAWs, super-high jpgs or transferred tifs in comparison to lower jpgs that were processed and compressed in-camera. No one's going to match them up without a LOT of work - period.