Yet another example of R1 Raw vs. Jpeg

[mediokre]

Taken simultaneously. One RAW, one 10MB JPEG (high quality, sharpening -1, all other settings default).

I regret the poor presentation, particularly the inconsistent cropping and my failure to attend to the contrast of the white roof. I should've given it more thought. I'd really just wanted to show how the texture in the purple building is lost in the JPEG, and how this loss was not due to insufficient sharpening (hence the 3rd, obviously unacceptable JPEG sample). I didn't intend this to be a display of my postprocessing skills or a photo exhibition. I guess I deserve to be roasted.

 

[dkny]

That must be why RAW images use more memory and take longer to write to the card. Radical. Who would have thought that?

But if you're not printing larger than 8x10, JPEG will suit you fine. Really.

 

[Joel]

I'd seen another dramatic illustration of R1's Raw-Jpeg difference
(a white house), but I'd been skeptical, especially after some
people on this forum insisted there was no difference. Having taken
this photo myself, I was finally convinced me that R1 Raw really
has much better IQ. I am sorry it took me so long to see the
obvious.

It depends a bit on how you rate IQ and the processing you use on RAW and JPEG, which require quite different treatment and workflows.

The RAW file here has higher contrast so looks sharper at the edges and some of the features look more like noise than revealed detail. Despite the harsh PP the JPEG has undergone, it has better tonal range in the details (like the smears on the walls).

Of the images you provided, personally, I prefer the JPEG despite my liking for clinically sharp images. I would actually soften the RAW image a little and sharpen much more sensitively with a selective tool like FocalBlade and not the levels of USM used here.

Also, sharpening for print and for screen are different exercises with different optimisations. For normal printing, the RAW would probably appear slightly sharper as it's far more aggressive but if the print were poster size it might not look so great.

My feeling is that, with effort, you can coax a little more from the R1 RAW in terms of detail but it's not dramatic and it's not always actually that attractive, given what you have to do to the image to reveal it.

Joel.

 

[David A Melges]

...pretty much exactly my take.

dave
--
Amazing what we can do with just three crayons, red green and blue!
http://diamondmultimediagroup.com

 

[inframan]

...gobbledegook to me. So what is this the new anti-RAW rant?
Any post is better than no post?
--
Shoot first, analyze later!

http://www.flickr.com/photos/inframan

 

[David A Melges]

...I went back and did a 3rd objective comparison of RAW to jpeg.

First I should say the scrutinizing at 1 to 1 is not my preferred method of making a decision...real world output is all that really matters to me, and in this test, exactly as in the first two, the prints, simulating 20x30 inch prints there is zero observable difference for me, and anyone I showed the prints to.

For this test though I went to the step so many others seem to like, the one to one screen test.

I shot a magazine cover...I prefer this because with text on a solid background you get very clean detail. I tripod'd the camera, used ISO 160, preset white balance, full 10mp fine jpeg with RAW in one shot. All other settings are default. I used the timer and F11 and therefore have an exceptionally steady, sharp image.

I output the RAW to a tiff, cropped the two shots identically and here they are above and below:



To my eye its obvious which is the jpeg, its slightly softer. Just as Joel pointed out above however, its not so much detail that appears to be missing as unrevealed noise.

The real problem is that this is NOT 1 to 1....its 2 to 1. Thats where I had to go to really see a difference. Definitely nothing as dramatic as the original poster. Not sure why I never seem to find the big difference that others do...I only know that I don't.

In my opinion only, shooting RAW for anything but very demanding llighting situations is a waste of card space and pp time.

Even enormous prints from jpeg, if properly exposed in the first place look exceptioinal, which is the real test for me.

Just my take.

dave
--
Amazing what we can do with just three crayons, red green and blue!
http://diamondmultimediagroup.com

 

[inframan]

--
Shoot first, analyze later!

http://www.flickr.com/photos/inframan

 

[mediokre]

Thanks Joel for the information and education.

I should've just said "details" instead of IQ, which is rather subjective. I just used the default settings PhotoShop chose for me when I opened the RAW, as my main concern was with the texture of the purple building in the middle. I am quite sure that you can't coax it back from the JPEG. That's all I was trying to say.

 

[JortS29]

However, as in general you don't always look at a photo on a 100%
crop and many people do not even print larger than 7x5 than, if you
know how to shoot right, most people won't see much of a difference.

True, but what you seem to forget is that alot (if not all) people crop pictures from time to time. Especially with a camera like the R1 that gives you the equiv. of 120mm at the tele end, those objects out of reach can be brought closer with cropping. Any extra detail at 100% viewing is helpful in that area.

The non sharpened jpg crop posted later show
clearly that the lack of details is due to bad focusing which in
this case, explains the difference in details between the two shots.

Clearly to you. If you read other replies from the OP, you will notice both shots were taken at the same time. The jpeg came with the RAW, only the cropping differs...l.

On the other hand, the roof portion on the right, is better
detailed in the jpg version than in the raw one. The reason for
this is blown highlights in the raw version. What conclusion can
you draw out of this then?

Not much, apart from the fact that the PP on the RAW blew out the highlights. Looking at the jpeg, we can conclude that the RAW does contain atleast as much information on that part. Detail that was lost in the jpeg on the other hand, cannot be retrieved.