Scheelings
Well-known
The jpeg's coming out of this camera are generally lovely, however after one particular shoot with mixed lighting I got mixed results OOC.
After that I decided to switch to RAW + Jpeg with the view that the additional latitude a RAW file offers would give me an advantage when correcting white balance etc.
After taking some shots however, I discovered that for some reason the RAW files were much noisier. Has anything been written about this before? Is there a guide?
After that I decided to switch to RAW + Jpeg with the view that the additional latitude a RAW file offers would give me an advantage when correcting white balance etc.
After taking some shots however, I discovered that for some reason the RAW files were much noisier. Has anything been written about this before? Is there a guide?
tbhv55
Well-known
My guess would be that you have the Dynamic Range setting on something other than DR100 (DR100 equates to OFF). If this setting is not on DR100, the camera is then at liberty to use higher ISO values when trying to 'balance out' dynamic range.
There is a brief (and not very helpful) reference to this on page 70 of the User Manual. Alternatively, Googling "Fuji x100 dynamic range" will offer a number of discussions on this, in the first few hits.
BLKRCAT
75% Film
Are you shooting full auto or anything auto? Do you have noise reduction set on your jpgs? What version of x100? How are you viewing the images? Have you checked the meta data to see what the images actually are being shot at?
The x100 shouldn't be shooting two simultaneous images at different isos. It will shoot the raw and in camera process the jpg, then depending on your settings keep the raw or delete it.
You're probably seeing what your images REALLY look like without the in camera voodoo happening.
Personally when I tried the x100 I found the in camera noise reduction to be terrible. Pretty much the case with all in camera noise reduction. Just loses detail, and turns your image into a pastel mess at 100%
The x100 shouldn't be shooting two simultaneous images at different isos. It will shoot the raw and in camera process the jpg, then depending on your settings keep the raw or delete it.
You're probably seeing what your images REALLY look like without the in camera voodoo happening.
Personally when I tried the x100 I found the in camera noise reduction to be terrible. Pretty much the case with all in camera noise reduction. Just loses detail, and turns your image into a pastel mess at 100%
willie_901
Veteran
I concur with tbhv55.
When DR =100 at ISO 1600 and below, there is no reason for the raw files to be "much noisier" than the JPEGs.
Above ISO 1600 (which I rarely used on the original X100) it is likely FUJIFILM applies in-camera noise filtering to the JPEGs. The noise filtering, if present, would be in addition to the in-camera JPEG noise filter parameter. However it is also possible the raw files are similarly filtered in-camera as well.
At all ISOs a JPEG with the highest, user selectable, in-camera, noise-filter setting may appear to have a higher signal-to-nose ratio than a raw file rendered with no noise filter.
As ISO increases the exposure (signal level) decreases. The signal-to-noise ratio decreases. ISO amplification only increases the rendering brightness, not the exposure (signal level). The amplification increases the signal and noise levels by similar amounts. The difference between an in-camera JPEG with the maximum noise filter and a raw rendering with no noise filter will be most noticeable at lower signal-to-noise ratios (higher ISOs, less exposure). However using the an appropriate noise filter during raw rendering should result in images with similar signal-to-noise ratios.
As far as I know there is no in-camera raw file noise filtering at ISO 1600 and below.
You might compare results with DR=100.
In identical light take a set of JPEGs at ISO 1600 (or the highest ISO you plan to use) with all the in-camera JPEG noise filtering levels.
Then take two raw files. Expose one at ISO 1600 (or the highest ISO you plan to use). Expose the second at ISO 200. The second will initially appear very dark. In post-production increase the raw ISO 200 image's rendering brightness globally by 3 stops. Then compare which image has the best signal-to-noise ratio in the shadow regions.
Of course you would want to compare the raw to JPEG as well.
For the original X100 I decided ISO 1600 was the emergency maximum for color work. ISO 1600 was acceptable for B&W. I only used raw files and rarely exceeded ISO 800.
When DR =100 at ISO 1600 and below, there is no reason for the raw files to be "much noisier" than the JPEGs.
Above ISO 1600 (which I rarely used on the original X100) it is likely FUJIFILM applies in-camera noise filtering to the JPEGs. The noise filtering, if present, would be in addition to the in-camera JPEG noise filter parameter. However it is also possible the raw files are similarly filtered in-camera as well.
At all ISOs a JPEG with the highest, user selectable, in-camera, noise-filter setting may appear to have a higher signal-to-nose ratio than a raw file rendered with no noise filter.
As ISO increases the exposure (signal level) decreases. The signal-to-noise ratio decreases. ISO amplification only increases the rendering brightness, not the exposure (signal level). The amplification increases the signal and noise levels by similar amounts. The difference between an in-camera JPEG with the maximum noise filter and a raw rendering with no noise filter will be most noticeable at lower signal-to-noise ratios (higher ISOs, less exposure). However using the an appropriate noise filter during raw rendering should result in images with similar signal-to-noise ratios.
As far as I know there is no in-camera raw file noise filtering at ISO 1600 and below.
You might compare results with DR=100.
In identical light take a set of JPEGs at ISO 1600 (or the highest ISO you plan to use) with all the in-camera JPEG noise filtering levels.
Then take two raw files. Expose one at ISO 1600 (or the highest ISO you plan to use). Expose the second at ISO 200. The second will initially appear very dark. In post-production increase the raw ISO 200 image's rendering brightness globally by 3 stops. Then compare which image has the best signal-to-noise ratio in the shadow regions.
Of course you would want to compare the raw to JPEG as well.
For the original X100 I decided ISO 1600 was the emergency maximum for color work. ISO 1600 was acceptable for B&W. I only used raw files and rarely exceeded ISO 800.
Scheelings
Well-known
Some useful tips here. I think I need to go back and look at the files in LR and provide some more data.
It's the original X100
It's the original X100