Fuji S5 Pro in 2018... Madness?

John Motts wrote:

EE-TV wrote:

John Motts wrote:

fPrime wrote:

John Motts wrote:

chris goes live wrote:

Thanks for that, I am now up and running with Fuji’s hyper utility hs-v3 . It certainly is a bit clunky, but it beats capture one for iq, which was a shocker. See here

Capture one version

https://www.flickr.com/photos/135677573@N04/33808850508/in/dateposted-public/

compaired to the hyper utility version

https://www.flickr.com/photos/135677573@N04/46770316065/in/dateposted-public/

They're just different settings.

Sorry but that’s wrong. I am fluent in C1 as well as Lightroom and Hyper Utility HS-V3. The difference in this case exceeds merely different “settings” as a native color profile and native demosaic is available to Hyper Utility.

fPrime

I'm fluent too - I do it for a living. I process for other people too.
You can't just say that one software is better than another with one single image straight off like that. No raw processing software is so different from another. All need different treatment but all can be made to look much like each other.

Sorry, but they just aren't that different.

Only Fuji's own supplied Hyper Utility is able to bring out the color rendition that one is to acquire via the camera's in-built processed jpeg. The camera is notorious for its film simulated colors and natural skin tones. So much so that it is one of the rare cameras that often the jpeg trumps the RAW...

https://www.dpreview.com/forums/thread/4390246?page=2#forum-post-62634887

Of those two examples, one is contrasty, one is muddy. It takes seconds to make the Capture One version look like the Hyper Utility version. With respect, one piece of processing software can't be dismissed on the basis on this overly simplistic comparison.

I have just spent a huge amount of time this winter while my weddings are less busy comparing RAW processing software, from Lightroom, Fuji's Silkypix, Fujifilm X Raw Studio (which is basically the same as the camera's own processing) to Capture One.
The overwhelming outcome is that the image quality differences between all of them are hugely exaggerated. Some need less work than others but used properly there isn't a vast amount in it quality-wise. Certainly I wouldn't dismiss one on the basis of one basically unchanged example.

Bottom of the field for image quality was Lightroom, but again only by a small amount. More tweeking was generally needed to get it there and the difference is nothing remotely like the amount implied by the two Flickr examples given above.

Next to go was Fuji's X Raw Studio, which uses the camera's own processor, mainly due to its highlight performance. Any of the other three can get significantly more from highlights. The other two were close but C1 won it overall in terms of of image quality and practicality / speed.

I should add that these were with X-Pro2 files.

I'm not dismissing any of these, just saying that no RAW processor does such a dreadful job as implied by the examples shown.

What you say is probably true for an xpro file (a camera I love by the way). However the s5 has these dual pixel sensor which creates a demosacing issue. I have found that capture one is not able to reproduce the detail of hyper utility. Look at the two images at 100% there is clearly more information (detail) in the hyper utility image. As for the more muddy look. I think this is due to the raw being flat, ie it collapses the dr to the 255 luminance range when exporting to jpeg. As I mentioned on Flickr the two images were produced using the auto setting in capture one and hyper utility, capture one goes for a punchy contrasty look, whilst hyper utility goes with a flat muddy look.

I now now have to go back through my catalogue of s5 images and see if hyper utility can improve on the raw out put.

many people have complained over the years on the inter web that the s5 has too strong an aa filter, and images are too soft. I now wonder if this is because most processing tools are unable to demosac the cameras output correctly?