Difference between an in camera jpeg and a raw file with Voigtlander 15mm, f/4.5 lens

[grebeman]

The following photographs are an unprocessed raw (converted to jpeg in Bibble Pro 5) and an unprocessed jpeg straight out of the camera, the only post processing being resizing in Bibble Pro 5.

Subscribe to see EXIF info for this image (if available)

original jpeg

Subscribe to see EXIF info for this image (if available)

original raw converted to jpeg

My attention was drawn to the upper right hand corner of the shot where in the out of camera jpeg there appears to be a very thin greyish cloud, however there was none in reality, it being a clear blue sky. The raw file hardly shows this effect at all, and in the original at original size it can't really be seen at all.
I saw the same effect in several other photographs taken whilst walking along the estuary. I was comparing this lens with the Panasonic 20mm, f/1.7 which exhibits no difference in output between it's in camera jpeg and the corresponding raw file.
Does anyone have any thoughts?

Camera used Panasonic GF1

Barrie

 

[OzRay]

There is obviously something there, even though you didn't see it yourself; very wide angle lenses can easily have this effect. Nor can you compare the JPG engine with RAW converters, unless you use the Olympus one, as programmers have to guess how to convert the RAW images, some do it well, others don't. Personally, I've never liked Bibble.

Cheers

Ray

 

[soundimageplus]

Its part of a phenonenum called "maze artefacts". See this link:-
Flickr: Discussing maze artefacts in Micro Four Thirds cameras using alternative & legacy lenses

Its apparently a demosaicing problem that exists with certain software raw development. Photoshop or Lightroom have the latest Panasonic patches, which should fix it or you can use the Silkypix software that came with your camera.

I had exactly the same thing with this lens. There are also Zeiss wide-angles which do the same thing.

 

[grebeman]

Ray,
I've seen an artifact generated by RawTherapee, interestingly enough in the same corner of the photograph when using the Voigtlander 15mm, f/4.5 lens which consisted of a feint square pattern, which at the time I thought might be due to my sharpening the photograph before resizing it to post on the web.
This is the first time I've taken raw and jpeg with this camera and lens, having been prompted to try jpegs following a recent thread on this forum. Had I just used raw as normal I would not have seen this.

Barrie

 

[silverbullet]

Hi Barrie,

interesting point you show us. Last year I had the same effect with a file from a Ricoh GRDIII (28mm in135) f1.9. It's the standard fixed lens and here is the pic with the gray bad-weather sky at the right side.

The original RAW file had a slightly burnt sky and in LR2 I pushed a little bit to pronounce the clouds and the blue of the sky. At the right side - nope - no blue but gray..... I take it as it is, but it would be interesting what it is....

here the link to flickr:

Old glory | Flickr - Photo Sharing!

 

[BillN]

Barrie

are you happy with the browns in that jpeg - I saw the same thing in your weather shot, and you discuss it

- maybe it's me?

Cheers

 

[grebeman]

Hi Barrie,

interesting point you show us. Last year I had the same effect with a file from a Ricoh GRDIII (28mm in135) f1.9. It's the standard fixed lens and here is the pic with the gray bad-weather sky at the right side.

The original RAW file had a slightly burnt sky and in LR2 I pushed a little bit to pronounce the clouds and the blue of the sky. At the right side - nope - no blue but gray..... I take it as it is, but it would be interesting what it is....

here the link to flickr:

Old glory | Flickr - Photo Sharing!

Hi Bernd

I have taken onboard the thoughts expressed earlier in the thread that this is related to maze artefacts, and here is the photograph I posted some months ago that I would suggest demonstrates maze artefacts as I understand them from the above links, a definite pattern on the photograph

Subscribe to see EXIF info for this image (if available)

However what you have shown is similar to my photograph, and that is much more of a grey wash. When magnified it does show the very faintest of pattern, but it is not pixelated as is the rest of the photograph at that extreme magnification since it is a smooth tone.
Also soundimage states in his flickr discussion that the jpeg obtained from the camera was clear, in my case the jpeg obtained from the camera was the most affected, and in your case you were using the native lens, so perhaps there is more to explain here than we currently understand, or have I missed some aspect of what constitutes maze artefacts?

Barrie

 

[grebeman]

Barrie
are you happy with the browns in that jpeg - I saw the same thing in your weather shot, and you discuss it
- maybe it's me?
Cheers

I will use the same argument, the local geology will have an influence on the shade of soils, estuary silts and muds, I am very familiar with that estuary, if it wasn't for one rising field between the estuary and my house I would be able to see it from my house. I walk it frequently bird watching and I have no problem in accepting the colour rendering of the brown in that photograph, I'm assuming you are referring to the bulk of the foreground. It will change with how it catches the light, how much it dries out between tides and possibly other variables.
In what way do you feel it doesn't represent the truth?

Barrie

 

[BillN]

I prefer the colour in the RAW image

both banks plus the wall

 

[grebeman]

I prefer the colour in the RAW image

both banks plus the wall

Bill,

I would agree. I normally shoot in raw, have been shooting raw and jpeg recently, my interest being aroused by the following thread https://www.mu-43.com/f35/return-jpeg-6580/. I'm still tweaking the settings in white balance, contrast and saturation to see how the in camera jpegs come out. I need to have a nice day weather wise when I can set up on a tripod and alter settings, making copious notes and see what gives me the best results.
Hopefully I'll be able to post here at some time, but with winter upon us, well I can't promise when :smile:

Barrie

 

[silverbullet]

the JPEG pic is more saturated and it seems for me that there is some sky reflection (blue) on the wet mud.

Regarding the Ricoh pic I have seen somewhere a comment about this uncolored gray hue.

Sidenote: in the Leica forum LUF two years ago Sean Reid, the guru of photography, tested wide angle lenses from Leica and Voigtlaender together with the Pana G1 when I remember well. He pointed out that the lenses from 35mm and shorter would cause problems at the edges. I don't know whether he only mentioned resolution and/or color shifts due to interaction of coverglass on sensor etc..
It's funny because the sensor has such a huge crop factor.....

 

[grebeman]

Bernd,
Apart from the mesh type artefact in the photograph posted above which at the time I thought was due to sharpening before reducing to post on the web, but now I think it was due to it being post processed in RawTherapee, I've never noticed any problems with the 15mm Voigtlander lens. It's only now that I have used it to produce in camera jpegs and perhaps one of the few occasions when I have photographed clear blue sky that I see this effect.

Barrie

 

[silverbullet]

Barrie,

just now I found a statement about blown highlights in sky areas. Interesting fact about several blown color channels:

"When you clip one or more channels, you loose the ability to generate accurate color so Camera Raw tries to give you textural detail even if the color won't be accurate. It actually works well when only one channel is clipped...it works less good when you have two channels clipped. Many converters actually quite trying to get image data past the first blown channel. You can do into the Adjustment Brush and paint in some color over the recovered texture...that's about the best you can expect. Once channels are blown, Camera Raw would have no idea what colors those blown channels WOULD have produced if they weren't blown so it'll be up to you to add the correct color."

Could be true in my case but will not help with your effect.

 

[grebeman]

Bernd,

I've just looked at the raw photograph of the boat and the bridge in Bibble Pro 5 and pushed the saturation up to 100%, but that doesn't cause any grey degradation of the sky in the top right hand corner as seen on the in camera jpeg. Moral-don't use the jpegs from the 15mm Voigtlander, however raws seem unaffected, hence my satisfaction with the lens up until now. I think it will remain my favourite wide angle lens even if the 20mm Panasonic produces sharper results, I can't break that habit of using those lovely engraved markings on the lens barrels of old lenses, even if in :43: some mental arithmetic is required to use the depth of field markings.

Barrie

 

[Ulfric M Douglas]

This is the first example I've seen of problems with full-frame wideangle lenses on m4/3rds ... and it looks like this is a minor thing. I guess that's good.
The colours on the RAW-Bibble-Jpeg (it isn't a RAW file, its a Jpeg made by Bibble instead of Panasonic.) are much nicer than the OOC Jpeg, so obviously either bibble is doing great work or Barrie has chosen the wrong film-mode for that particular scene.
which JPeg did you choose Barrie? (Personally I try to avoid Standard.)

 

[grebeman]

It probably was standard, I think the thread I referred to suggested dynamic from memory. I'll try some tests when I get the chance to set up a scene with a tripod mounted camera.
I take your point, you're not being shown the raw file, but a jpeg generated by Bibble. I have the advantage of being able to see the raw file and should have produced a jpeg to see what that would show.

Barrie

 

[grebeman]

Right, just produced a jpeg from the raw with 100% saturation, a little grey clouding in that corner, but nothing like as much as the in camera jpeg.The sun was off at an angle in that corner, perhaps therein lies the answer, some sort of flare effect, possible generated in the camera with light bouncing around inside, all that spare image circle.

Barrie

 

[pdh]

hmm ... there are those who say that there are problems when a lens of < 35mm fl is used on mFT ... wonder if this is the sort of thing they are meaning?

 

[grebeman]

Whatever, there is much less of an effect on a jpeg processed from the raw file than there is on a jpeg generated in the camera, at least when using Bibble Pro 5 to open the raw file and produce it's jpeg and the GF1 to generate the in camera jpeg, oh so many variables, my head is beginning to spin.

Barrie

 

[Brianetta]

I prefer the colour in the RAW image

It stopped being raw the moment Bibble cooked it.