GRDII vs GRD ISO test

[Drafi]

Sorry, Pavel, with respect to the upscaling, I obviously did not know exactly what I was talking about. It seems upscaling just means to format the picture so that it fits the size of the picture you are comparing it to.

[asabet]

Pavel, Thanks for providing these samples! As you may have seen, I processed some crops from your RAWs and used them to start a discussion over in the DPR forum as well as on my blog. I assumed that by sharing the RAWs for download you would be okay with this, but if you want me to pull them down, just let me know. I of course credited you and this site for the files.

One thing that concerns me is that the GRD2 high ISO noise seems to me to have a "blotchy" quality. It makes me wonder whether Ricoh is applying more NR (and sharpening) to the RAW files with the GRD2.

For example, I took crops from the areas highlighted in yellow:


All the following crops are from RAW and processed with NR and sharpening settings at the minimum amount possible with the designated RAW processing application. No resizing.





Regards,
Amin

[Gil]

Amin- Is it possible that the difference that you notice is just due to the way the new incamera prossesing engine
interprets pixels instead of increased NR and incamera sharpening?

Gil

[dasper]

Gil - ??? Those are precisely functions of the processing engine.

[Gil]

Yes, you are obviously correct. I did not really pose the question to well nor was my thinking very clear at the time I guess.
Sorry,
Gil

[mr.datsun]

Guys, you can get the original (unresized) photos from the initial post. Upscaling the smaller photos is a widely used method (although probably not the best) to present similarly sized results from two differently sized sources.

Thanks - I didn't realise that the files you included were unresized. Even the unresized GRD I jpegs look quite softer - could this be due to any difference in in-camera image settings?

I would still like to ask what others think of the unusual pattern that shows up in the 1600 ISO GRD 2 JPEG and also to a lesser degree in the 800 JPEG. For an image with no noise reduction they're distinct digital artefacts and not something I can see in the RAW files..

[Studio67]

Many thanks for your work!
I have the DNG compared (800 iso) and I reached the conclusion that the GRD II is a lot better! Higher resolution and better image quality. There is a lot of nonsense written in the DP review forum about noise reduction in the DNG, but this is not true.

But a noise reduction function in such a small device can of course never equaled with a piece of software like Noise Ninja.

The Ricoh is a camera with its own identity and character.
We are talking about a camera between € 500 - € 700 costs!
Buy an M8 with all the problems and bugs.... € 6000

[lucridders]

how can I have a neutral view about the GRDII ? Is impossible with the tests we see now. They are always taken by changing the settings, using PP software and using Noise Ninja or similar.

I would like to see once pictures made in color from landscapes, still live, portrait and this with all settings to 0 and in auto mode. No post processing and no Noise reduction. Strange is also that I hear all the time that Ricoh cameras are no producing more noise than others, but on the other hand i only read posts and see images that have some postprocessing and noise elimination. My brains are really too small to understand this. So come up with pictures as they are in neutral settings and auto mode. Also pictures, with quite lot of difference in dark areas and highlight areas. This will tells more about a cam than all the rest together. When I like to do later on setings, I will do it myselves. In automode however, a cam should be able to capture for 75% right out of the box good images.

[odklizec]

I'm sorry Lucridders, but if you don't like these tests, you can go to flickr or another public photo gallery and search for tests which suits your needs better. I'm not a pro reviewer and the test I'm doing are the test, which interest the majority. Test with Noise Ninja is there to show how good can be eliminated the noise from RAW with preserving reasonable amount of details and pleasant to look grain. This test also shows the difference between such processed RAW and in-camera noise reduced JPEG, which, as I believe, is pretty evident.

And please, there is no need to post the same in multiple topics. Thank you!

[asabet]

Lucridders, I think you can count on a number of review sites, such as DC Watch and others, to give you exactly what you are asking for. Pavel is giving great information here, particular for those of use who want to know what the GRD II sensor is directly capable of delivering.