Jump to content

Snow on the JPG but not on the DNG??


Alberti

Recommended Posts

Advertisement (gone after registration)

This week I took some pictures in a light snow drizzle in Köln.

Strangely the snow shows up in the JPG-in-camera, but not on the DNG, while the DNG should provide more details and sharpness isn't it? You can clearly see the building is sharper in the DNG.

 

First the two overall pictures with first the DNG then the JPG, that was taken a few seconds later.

After that a details. (Time difference is 20 sec's. So I don't expect that to be the reason - that it started to snow all of a sudden.)

 

Am I seeing strange things? I noticed in another picture that the C1 might hide such small details.

Alberti

 

DNG-'developer': Capture One 4.8.3

M8, Summicron 50, at 5.6, 250 and at 360 sec.

Link to post
Share on other sites

If you are saying that the snow is more visible in the LCD image on the back of the camera than when you view the DNG with a RAW processor then I think the reason is as follows.

 

I suspect the jpeg image used for the LCD display is produced by under sampling the pixels to reduce processing time. In other words not all pixels are used to produce the jpeg. If the resulting jpeg is say 400 pixels in total and a snow flake happens to be sampled it will occupy 1/400th of the image. Compare that with the DNG where the snow flake may cover only 100 pixels in 10 million, that is 1/100,000 of the image area so the snow flake is much less visible in the DNG image.

 

In a similar manner under sampling also causes the camera's LCD image to appear more noisy at high ISO than the developed DNG.

 

I think if you set the camera to record a hi res jpeg and DNG at the same time there will be little difference in the snow visibility when you view the hi res jpeg and DNG.

 

Bob.

Link to post
Share on other sites

Alberti--

I don't think it's a good idea to try to draw conclusions from two different images made at different times. Better would be to set the camera to record DNG and JPG simultaneously. That's the only way to get a one-to-one comparison.

 

You're right, twenty seconds isn't a long time between the shots, but a gust of wind could have driven the snow back in the DNG shot, or the snow might simply have been lighter then. I understand the question and the concern, but if we're trying to compare DNG and JPG, the only way to make a meaningful judgment is to compare two identical images.

 

There's certainly a difference between these two, that's for sure! :(

Link to post
Share on other sites

Advertisement (gone after registration)

Alberti--

I don't think it's a good idea to try to draw conclusions from two different images made at different times. Better would be to set the camera to record DNG and JPG simultaneously. That's the only way to get a one-to-one comparison.

 

You're right, twenty seconds isn't a long time between the shots, but a gust of wind could have driven the snow back in the DNG shot, or the snow might simply have been lighter then. I understand the question and the concern, but if we're trying to compare DNG and JPG, the only way to make a meaningful judgment is to compare two identical images.

 

There's certainly a difference between these two, that's for sure! :(

 

HoCo, I agree that it is not scientific.

 

Now I did see (now again) an interesting phenomenon on opening the JPG in C1. I could see the flakes disappear : as the image refined and got recalculated, they 'melted'.

 

Adobe says about the RAW capture and conversion:

Noise reduction, antialiasing, and sharpening. Problems can arise with very small details in an image. If the detail is only captured on a red-sensing pixel or a blue-sensing pixel, the raw converter may have a hard time figuring out what color that pixel should really be. Simple de-mosaicing methods also don’t do a great job of maintaining edge detail, so most raw converters also perform some combination of edge-detection and antialiasing to avoid color artifacts, noise reduction, and sharpening.

 

Just for sure, I like C1 in the high quality rendering of images. But,

 

My conclusion: C1 uses an algorithm, even on the on-screen display

  • that looks for straight edges
  • to the expense of discontinuities
  • so, spurious, one-off objects like a red berry, a snowflake ARE made invisible.

While Bruce Fraser says that a 8-by-8-pixel block forms the foundation of all normal JPEG compression, I would expect a more sophisticated treatment when working with the RAW image.

 

alberti (and this by the way does distort my vision of reality :rolleyes:)

Link to post
Share on other sites

Try using a DNG viewer instead of C1 which utilizes profiles and may likely be removing the snow thinking it's dust or dead pixels.

 

Long term, update or find another raw processor that retains or allows you to retain all image data.

Link to post
Share on other sites

Alberti--

Thanks for the info and the thread. Fascinating that you see the correction taking place in C1.

 

My earlier comment was rote and oversimplified; you've now brought me up to where you started. Thanks. :o

Link to post
Share on other sites

Honestly, there are a few flakes visible in the DNG--see above the sign with the arrows, and on the numbers "4711" on the large bottle-shaped sign. I think there just wasn't as much snow as when you snapped the jpeg...honestly, if the converter were erasing snowflakes, it would also erase the white spot on the pavement and the black dots on the glass doors. You are overthinking this.

Link to post
Share on other sites

Honestly, there are a few flakes visible in the DNG--see above the sign with the arrows, and on the numbers "4711" on the large bottle-shaped sign. I think there just wasn't as much snow as when you snapped the jpeg...honestly, if the converter were erasing snowflakes, it would also erase the white spot on the pavement and the black dots on the glass doors. You are overthinking this.

 

I'll keep a mind to it, though.

Really the other day I took a picture with red berries on a tree, that showed up on loading but were black when the picture was built up. I checked by pixel-diving (400%), the berries were also just a pixel or one two large.

alberti

Link to post
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...