On the left is a vlog jpeg image that I took on my Lumix S1RII. It has been put into davinci resolve and a colour space transform using panasonic V-log has been used. This is how the image looked when I took the photo. The image on the right is the same image but the RW2 raw file. What DxO is giving me is wrong even though I’ve set it to no colour corrections, and I can’t find anything that turned on that would affect this.
Can anyone help me because I don’t know what it’s doing.
The JPG has been edited in the camera. Look for something like picture control.
In PL you have to do the edit yourself, as in any raw converter.
George
Sorry I should’ve clarified that the image I put into PL was a raw RW2 file, so there’s not been any camera edits. I’ve applied the DxO modules which I thought was the colour space transform equivalent, but even removing them does nothing.
JPEG files are processed in the camera. Then you did further processing.
The RW2 file is totally unprocessed, even in the camera. If you intend to compare software, it should be with the same file type.
That’s not what’s happening. The jpegs are in Vlog so any colouring applied to them would match on a RAW file (Minus the flexibility of RAW). Basically my question is why is the RAW image not being processed correctly. Even with the zero corrections preset selected, it is still giving me an incorrect image.
Take the raw file into DaVinci Resolve and see what it looks like. If I am not mistaken, V-Log is for video only and not stills, so I would suggest the JPG is the colour space you selected in camera (probably sRGB or AdobeRGB).
Post the RAW and JPEG files here or send them via a link to WeTransfer
A simple Google. V-log is editing the image.
https://www.panasonic.com/uk/consumer/cameras-camcorders/lumix-expert-advice-learn/technique-technology/what-is-v-log.html
V-Log is just another Photo Style, but one designed to produce an image that has flat contrast and low colour saturation. All Photo Styles alter the way the image from the sensor is presented to us, adding contrast and colour characteristics to create a specific look or to match what we expect from certain subject matter. The Landscape/Scenery Photo Style creates deeper shadows and more colour saturation in yellows, reds and blue skies, for example, while Natural tries to create a look that matches what our eyes see – moderate contrast and moderate colour saturation.
It’s like the Nikon picture control I mentioned before. Bold is mine.
George
Searching further there is general consensus from Lumix users that V-Log is for video but can be used for photos. V-Log is not as good as raw and users say you should use raw due to far better dynamic range because of a larger bit depth. Users also say V-Log can cause major issues with photos.
@KeithRJ That doesn’t seem to be Panasonic’s opinion as discussed in the article that @George referenced.
I agree that it is usually used when shooting video. Our youngest son shoots and edits video for a living and has used members of the Panasonic GH gamily (4,5 and 6) and recently added an S5mkii to his inventory.
Apparently my G9 with also shoot VlogL.
I checked with my son and he said he never changes his settings from the Vlog setting he uses for video if and when he takes any photos.
@PerryB So in theory I should be able to take a VLog image with my G9 and then play with it in PhotoLab to see if I can reproduce your problem but it would be easier if you could supply the image and the DOP so we can actually investigate your actual problem.
Perhaps he should check Re: Lumix S5 - different video/photo settings: L-mount (Panasonic/Sigma/Leica) Talk Forum: Digital Photography Review .
VLog mode on Panasonics underexposes RAWs even more than Active D-Lighting on Nikons at ADL Normal setting. PhotoLab doesn’t take into account these settings in metadata, so you have to correct the exposure manually in PL. Obvious disadvantage of shooting in VLog photo or ADL modes is that you get more noisy photos.
It’s something as exposing for the high lights and correcting the shadows afterwards. In-camera you have to shoot in jpg for the correction is done on the rgb image. In a non native converter it has to be done manual.
This site contains 2 pictures with the same effect as @PerryB.
George
That is one reason why I have never used Active D-Lighting on my Nikon. It is one of the few settings that affects the RAW file.
And that was a very interesting article that clearly explains something I knew was “wrong” but didn’t know why.
If straight out of the camera this means that probably the exposures was changed and that that was corrected local in post processing, plus the other camera settings in your ‘photo style’.
@Joanna , I never used ADL either.
George
For some of my old shots taken with ADL enabled, I just correct the exposure and use SmartLighting in PhotoLab, to get similar results as in Capture NX2 or NX Studio (perhaps even with better color reproduction in PL). There’s more noise but ADL is for low ISO use.
Maybe Panasonic VLog is more like Nikon HLG (e.g. in Z8), though? But it still underexposes the RAWs, even by two stops, as some say.
Nikon ADL was meant to be used with Matrix Metering more or less to meter for highlights. I’ve used it sometimes in bright daylight at base ISO. It worked quite nicely with Capture NX2 but I later switched to using manual Exposure Correction if I wanted to preserve more highlights, like white clouds structure or just used point metering and exposure locking. This all depends on particular camera/firmware, e.g. I can see big difference in metering between D700/D4 and Z8.
When I remember well it forced the matrix metering for just that photo. With matrix metering the resulting exposure is determined by the 5 different surfaces. Based on an assumption it gives more priority to certain of these surfaces. By example with the image of a beach and horizon in the middle it tends to expose for the beach. That’s where mostly the subject is. With ADL it exposed to the sky and corrects the underexposed beach in post processing.
That’s what I understood. I don’t know about HLG.
George
It looks like Panasonic V-Log is more like Nikon N-log (compared both specs). Nikon uses it for video only, and rightly so. Don’t know why Panasonic didn’t disable V-Log style for photos (exiftool value 17 for PhotoStyle 0x0089 tag, which very probably PL doesn’t take into account), as it doesn’t look to make any sense using V-Log for stills. Except perhaps, if you want to be safer with highlight recovery with matrix metering, like with Nikon ADL, but then you have to manually adjust the exposure in post-processing (unless your software understands fully Panasonic makernotes, which still might be not enough). The shadows may get posterized if lifted up, which is not a big problem for video, but for photos it certainly is. Some Oscar winning movies have a lot of posterization and noise in night scenes but jury didn’t care about it. Maybe what video people mean by shadow recovery is adding some gray background, like starting the tone curve at x=0 with non-zero ‘y’ value? With photos, it’s a totally different story, unless you do some sort of B&W.
I had a look in PL8 at two RW2 raws taken with DC-S1H and V-Log photo style, which I found in the Internet. Metadata looked a bit strange, no correct lens info was there (things like aperture=0, or no lens attached), so PL displayed a warning icon on the thumbnails and disabled Lens Sharpness Optimization and Auto modes in Distortion and Vignetting, but that’s not important here. Raw data histograms didn’t show any missing values above black level (512) in the low range, which you would get if logarithmic-like digital curve was applied. It seems that the raw data was linear with respect to incoming light, so no “log” curve was digitally applied. The raw data looked underexposed by almost two stops (depending on the criteria). Maybe some video cameras have non-linear circuitry to provide “logarithmic” output at the analogue stage (??), but the camera in question has Sony Semiconductors IMX-366 sensor, which has ADC on-board. To sum up: RAW data was linear and underexposed by nearly 2 EV (Nikon ADL does it usually in -1/3 – -1 EV range). BTW, maybe there’s some problem in PL with BlackLevel subtraction for some Panasonic settings, but I don’t have the equipment to check.
BTW, something like a “log” curve (or inverted gamma like) is used by many cameras makers for lossy raw compression. See for example NEF Compression by Bill Claff for precise description for Nikon, probably similar to Canon, Sony, … Let me cite two observations by Bill (NEF means Nikon RAW, translate to CR3, RAF, ARW, …):
You may not agree with point #1 above for some types of portraits, but then you typically underexpose them anyway. Film was more forgiving…
