GR User Forum

The spot for all Ricoh GR camera users

Register for free, meet other Ricoh GR users, share your images, help others, have fun!

Tell your friends about us!

GXR; RAW v's JPEG 'out of the camera' Images

Wiener

Active Member
A question for those of you who know more about these things... ;)
I was asked take record photographs of a friends art exibition. Although the results came out OK working from the RAW images, I came across an unexpected phenomenon with the JPEGS.
In order to try and capture the 'true' colour and contrast I set my GXR up to take RAW plus a 'bracket' series (+/- 0.3) in both Standard and Natural settings.
However, all the JPEGs straight out of the camera were processed to a very low contrast and with dull colours. I attach one set as an example. The RAW's do look a little brighter than my 'eye' remembered them, but would one normaly expect such a difference between the RAW and JPEG files straight from the camera?
The examples posted here are only cropped a little; no contrast, sharpening, or other PP has been carried out on any of the files.
Note; the brighter image of each image is/was the RAW file.
Cheers!
Andy
 

Attachments

  • EXIF
    R0011708.jpg
    118.8 KB · Views: 1,132
  • EXIF
    R0011708-1.JPG
    316.1 KB · Views: 1,137
  • EXIF
    R0011709.jpg
    126.9 KB · Views: 1,129
  • EXIF
    R0011709-1.JPG
    320.6 KB · Views: 1,126
  • EXIF
    R0011710.jpg
    112.6 KB · Views: 1,121
  • EXIF
    R0011710-1.JPG
    323.2 KB · Views: 1,124
Hi Wiener,
I had a look at the histograms of all three JPEGs and found them to be centered and very narrow, leading to the dull look. As a start, I simply used autocontrast to spread the histograms over the full range, with led to much more contrasty samples (but not optimal ones, of course...).
Thus, fiddling around with levels in PSE (Photoshop Elements), its bigger brother or any image processing program should lead to better reseults.
Best regards,
Gerd

Original (small, with histogram)


After autofocus (with histogram)
 

Attachments

  • Bildschirmfoto 2011-01-17 um 19.27.27.png
    EXIF
    Bildschirmfoto 2011-01-17 um 19.27.27.png
    361 KB · Views: 1,478
  • Bildschirmfoto 2011-01-17 um 19.27.38.png
    EXIF
    Bildschirmfoto 2011-01-17 um 19.27.38.png
    485.1 KB · Views: 1,463
By the way, the RAW versions look too bright to me, as also the histograms indicate. But maybe, they fit the original impression...
Best regards, Gerd
 

Attachments

  • Bildschirmfoto 2011-01-17 um 19.37.39.png
    EXIF
    Bildschirmfoto 2011-01-17 um 19.37.39.png
    511.4 KB · Views: 1,456
Finally, this is the result if you move the right slider only in the "Levels" function - more or less what you get with the RAW...
Gerd
 

Attachments

  • Bildschirmfoto 2011-01-17 um 19.47.44.png
    EXIF
    Bildschirmfoto 2011-01-17 um 19.47.44.png
    454.2 KB · Views: 1,448
Hi Gerd, thanks very much for your input.
You are corrrect to say that the RAW files are too bright; this was my visual obseravation also.
My 'best guess' (yet to be confirmed by the artist) was as per the attached JPEG.

Would a 'too bright' RAW indicate over-exposure? The JPEG images, in contradiction, suggest under-exposure...
Thanks again,
Andy
 

Attachments

  • EXIF
    R0011709SP1s.jpg
    2.8 MB · Views: 1,050
I have no experience with RAW -- I am a pure JPEG shooter (and hobbyist...).
However, I would agree that the RAW images are overexposed (for unknown reason...). As the histograms of the JPEGs are "centered", I do not think that they are over- or underexposed. However, the camera should spread the histogram much more. Probably, this is only possible with some sort of DRO (dynamic range optimization?), as it was called on my Sony DSLRs. The DR function of the CX series would also help here (at least, I hope so...). I use a function called "Beleuchtung anpassen -> Tiefen/Lichter" in PSE for this purpose (or the central = gamma slider in the levels function -- or both).
All in all, for RAW, others have to jump in!
Best regards, Gerd
 
The sample you show also looks fairly dull. However, like the JPEGs it has a "centered" histogram, as the following screen shot shows:


Using auto-contrast, that is moving the left and right slider to the histogram (as one usually does), lead to a too contrasty image:


Here, I moved only the right slider towards the histogram - that looks better:


Best regards, Gerd
 

Attachments

  • Bildschirmfoto 2011-01-17 um 21.27.40.png
    EXIF
    Bildschirmfoto 2011-01-17 um 21.27.40.png
    441.3 KB · Views: 1,401
  • Bildschirmfoto 2011-01-17 um 21.28.00.png
    EXIF
    Bildschirmfoto 2011-01-17 um 21.28.00.png
    562.5 KB · Views: 1,388
  • Bildschirmfoto 2011-01-17 um 21.28.27.png
    EXIF
    Bildschirmfoto 2011-01-17 um 21.28.27.png
    484.1 KB · Views: 1,370
Gerd,
I have been using Silkypix Pro to work on the files, but hope to get hold of a copy of LR3 sometime soon, which I'm guessing will have similar controls to your 'Elements' software.

I too found it easy to 'improve' on the original using SP Pro. What I probably need to do is get a sight of the original again to check how close I am to the correct colour/contrast balance. The question as to why the RAW looks over-exposed (to bright) is perhaps the main mystery here. I'm guessing that the JPEG is indeed 0.5 Ev underexposed...I should take a grey card with me next time I try an exercise like this! ;)
Andy
 
I also find it hard to "recreate" the impression that I really had in my photos. Maybe, a gray card is the solution...
Gerd
 
Andy, what image settings you used?

The "overexposed" RAW is in my opinion a result of Ricoh's attempt to boost the sensor sensitivity. Something similar is seen in GRDIII RAW files. This also explains why the GRDIII or GXR S10/A12 are much more sensitive for light than their closest competitors equipped with the same sensors. See for example DPReview DR and ISO tests. Ricoh cameras delivers noticeably better DR and faster shutter speeds than the compared cameras.

The reason why the JPEG looks somewhat dull is in my opinion Ricoh's attempt to make them easier to edit? It's a very common thing in pro-level cameras that the JPEG files taken with standard image settings are often duller than for example out of camera JPEG from mass market cameras. Such files are much easier to edit without the risk of damaging details and visible editing artifacts. On the other hand, Ricoh approach is maybe a bit too conservative in case of A12 JPEG files? ;)
 
odklizec":3arssd4l said:
Andy, what image settings you used?

The "overexposed" RAW is in my opinion a result of Ricoh's attempt to boost the sensor sensitivity. Something similar is seen in GRDIII RAW files. This also explains why the GRDIII or GXR S10/A12 are much more sensitive for light than their closest competitors equipped with the same sensors. See for example DPReview DR and ISO tests. Ricoh cameras delivers noticeably better DR and faster shutter speeds than the compared cameras.

The reason why the JPEG looks somewhat dull is in my opinion Ricoh's attempt to make them easier to edit? It's a very common thing in pro-level cameras that the JPEG files taken with standard image settings are often duller than for example out of camera JPEG from mass market cameras. Such files are much easier to edit without the risk of damaging details and visible editing artifacts. On the other hand, Ricoh approach is maybe a bit too conservative in case of A12 JPEG files? ;)
Pavel,
camera settings were ISO 200, WCB to Standard + Natural (-0.3, 0.0, +0.3 Ev) and 'normal' Programme.
I guess I had expected the JPEG's to give me a range of colours and contrasts around which to process the RAW image. I have never before tried a challenge where the 'true' original had to be replicated, hence my puzzled look when I saw the results initially on the screen! ;)
Your insights into the Ricoh RAW boost as well as the 'conservative' approach to GXR JPEG processing engine makes everything much more understandable! :D
Perhaps I did not get is all so terribly wrong after all! :)
Clearly all is possible with 'post processing'. However, it does make one realise that those who stick to a doctrine of 'zero PP' and 'only images straight from the camera are authentic' are (as indeed in the days of film) IMHO, giving themselves a hard road to walk!
Cheers!
Andy
 
Thanks for a really stimulating discussion. I have been wondering about camera color accuracy and photo quality and the aesthetics of the finished image. All topics for a long discussion. For the most part I am interested in the image, not the accuracy. But Andy's set of photos for the artist brings up the question of color calibration as well as pp.I made a small trial of a picture in our house, and the result is 3 photos . The first is the jpeg as shot. The second is the raw as shot using just the embedded color information. The third is the raw adjusted to match the original. I brought the original into the same room as the display and adjusted the temperature (white balance) and tint under the direction of my wife who is an artist.My conclusion is that for my own purposes the raw (dng in LR-3) gives a very easy path to accuracy. I tried correcting the jpeg but found it quite impossible to match the original in color and detail. This experiment did not show the large light/dark difference that Andy obtained with the raw and jpeg and that I was expecting. I am not a purist who tries to get the image as shot to be the finished product. Sometimes it happens, but often it doesn't
The shots were taken today inside, cloudy, snow falling and wb set to cloudy.
 

Attachments

  • EXIF
    Painting raw corredted-1003531.jpg
    409.3 KB · Views: 1,047
  • EXIF
    Painting raw from camera-1003532.jpg
    421.7 KB · Views: 1,044
  • EXIF
    Painting jpeg from camera-1003532-2.jpg
    388.3 KB · Views: 1,047
I know the artist wants a different look...but even so the high contrast work looks amazing to me. Shows the brush strokes and the darker copper looks amazing to me.
It would look awesome in the office... ;)
 
Back
Top