digital artifact or simply reflection - RAW vs JPG

[3sh]

<p>Hi all, last weekend I was in Rutgers to meet my friend after a long time. When the motive of visit was purely a reunion type, but I could not stop myself using my camera whenever possible. I only attached the 50mm 1.8 with my D300 and used it all the time. So when we were going to the student center(walking by) for lunch, I took some random pictures. I had never before used the 50mm/1.8 in broad daylight, and I never compared my jpgs with the raw to find out what difference they produce. Actually since I started using RAW, I use the JPGs to select out the good and sharp pictures than the rest(mostly in every outing, I took at list 300-400 pictures with 12bit RAW+fine JPG, and later use the JPGs to find out the sharp and good from the rest. My laptop can read the JPGs very quickly, but takes time to open the RAW files, so I save some time by this way).<br>

So this time, I accidentally come across to look at the same JPG and RAW at 100% view one after another and notice a considerable difference in this picture. When I say considerable, I mean at 100% view, which no one might be interested to view at. And it is not showing any difference in structural view point rather in pattern of light, I believe producing from reflection.<br>

As I am not sure what is going on, I thought to ask for the answer. Being a physicist, I should be able to figure out the cause, but being an experimental physicist, I rely on hands on experiment and outcomes. As for the present time being I cannot repeat the process, I will rely on your replies to know the reason. //So here in the problem - look at the chimneys and you will see what light pattern I am talking about, is it digital artifact or reflection of light?// I am not sure, but from the curvature of the pattern, this is more likely to be reflection of light, but then why the JPG file removed it(it's still there but hardly visible) and the RAW contained it(I played with the RAW file with different settings, but cannot remove the pattern)! I will attach 3 pictures, the original, 100% crop from the RAW, and 100% crop from the JPG. The original is straight from RAW to JPG via adobe camera raw and then re-sized to fit the limit of PNet posting, nothing else is changed.<br>

Picture details - D300, f/4.5, 1/2500s, ISO 200, 50mm/1.8, metering-pattern, manual exposure, WB-AUTO at 10:03 AM, bright sunny day.<br>

Any explanation will be appreciated. Thanks.</p><div></div>

[3sh]

<p>Next one-</p><div></div>

[3sh]

<p>The last one -</p><div></div>

[bobcossar]

<p>Looks like a Newton's Ring, or other kind of interference, effect....but I have no idea why that would, or could happen.....Robert</p>

[lex_jenkins]

<p>Looks like moire from an interference pattern in the texture of the metal. I've heard of similar problems from other photographers but can't recall the specifics. I seem to recall they mentioned it was a quirk of some raw converters. Try a different raw converter and see if the problem disappears.</p>

[CoryAmmerman]

<p>The left side of the chimneys appear translucent, which leads me to believe that they are covered with a mesh-like material (possibly for heat transference). I've seen that before on household chimneys. This could be the root cause of your interference pattern. As for why the difference between JPEG and RAW, I'm sure it has something to do with differences in how the image was encoded, but I don't know anything about how that works and can't even begin to offer an explanation.</p>

[3sh]

<p>Lex, I think you are right, this is some kind of moire. If I ever get another chance to visit Rutgers, I will check it with the other side of the chimneys. I mean I will test along the sunny side and the opposite site, so that I can make a definite conclusion. I tried your suggestion with picasa, that's the other one I have at present, but the patters are still there. Attached is the 100% crop from picasa of the RAW file.</p><div></div>

[joe_c5]

<p>This is a guess, and short on details, but I suspect that the camera and the raw converter software are using different demosaicing algorithms to fill in the missing information from the Bayer pattern of the sensor.</p>

<p>Instead of every sensor pixel measuring red, green and blue, half will record only green, one quarter will record only red, and one quarter only blue. Since filling in the other colors requires making assumptions, different software will not necessarily come up with the same final output.</p>

<p>The raw converter software may try to retain more fine detail than the camera, at the risk of more artifacts.</p>

[Alan Marcus]

<p>

<p >The moiré is an artifact of an imaging system resulting from interference between two regularly spaced sets of lines. </p>

<p >You can observe a moiré pattern by peering through two layers of ordinary window screen. This business of making images is plagued by the moiré. You see it all the time on TV particularly when clothing has a pattern. The moiré is a clash or better described as a mismatch of the matrix of pixels. </p>

<p >Our cameras image using a chip built is a matrix pattern. We display on a screen composed of a matrix of picture elements. Our compression software acts on the matrix. The computer display is incapable of displaying the pixel density as recorded by the camera. The display adapter board with its chip logic casts out a high percentage of pixels. Prints on paper are comprised of a dot pattern and the various colored pixels are laid down at different angles. All this and more produce the moiré. </p>

 

<p > </p>

 

<p > </p>

</p>

<p > </p>

[matthew_rusbarsky]

<p>Disclaimer: I do not have a digital camera, only shoot B+W film, and know nothing about RAW. When PP neg scans for the web I usually end up with artifacts like these two ways: scaling an image or converting from a non-compressed format to JPG. When scaling an image I sometimes need to play with the algorithm to eliminate artifacts (sometimes cubic interpolation good, sinc bad, etc ). When converting to JPG I will sometimes need to go into the advanced options and play with the compression settings. Try converting your RAW to a non-compressed format like TIFF to narrow down what step in the process the artifacts are being produced. I might be that your camera is better and handling these settings than you are.</p>

[william l. palminteri]

<p>This is a Moiré pattern.</p>

<p>Bill P.</p>

[3sh]

<p>Thanks to all for their valuable suggestions. I think it's more like moire than a processing problem. Without even processing anything, 100% view of the RAW file is showing this pattern in any of the software I have used so far. I have saved the RAW in TIFF and still it is present. Certainly the RAW format is much more powerful than the camera converted JPG, so it is maintaining more information in it. I am getting closure into the physics of this moire occurrence and soon I will do some tests to draw a conclusion. Thank you all again for your valuable inputs.<br>

Hota</p>