 |
| The FZ300 is a very good street camera, well suited for quick photos of people and scenes as they present themselves. I have not shown the test photos referred to in this post. They are totally boring and reveal no information beyond that contained in the text. |
This subject comes up time and again on user forums. There are endless debates about the merits of high pixel count versus low pixel count sensors.
I quite often read confident statements by self appointed experts that “x (pick your number) pixels is simply too many for y (pick your size) sensor” as if this were some pre-ordained truth which everybody should acknowledge.
As it happens I have had an opportunity to test this.
I have on hand a Panasonic FZ300 (12Mpx) and a Panasonic TZ90 (20Mpx) each using the same size 4.55x6.17mm chip, presumably made by Sony although nobody is saying.
I photographed a test chart with each camera on tripod, OIS off, timer delay on at equivalent 50mm focal length and f4, (f4.1 on the TZ90) previously determined by me to give good sharpness with each camera, ISO 100.
I then photographed a set subject (books on shelves) with each camera at ISO settings 100-3200 again using a tripod, OIS off, timer delay.
I shot RAW and opened the files using Adobe Camera Raw. I set the sharpening amount at zero (default is 25) and luminance noise reduction at zero (default is zero).
I set the color noise reduction slider to 25. Setting this at zero produces a lot of color noise which confounds perception of the luminance noise. The color noise is easily removed completely from all the files so I do that for side by side comparisons.
I viewed the files side by side on a Dell ultrasharp monitor.
In order to meaningfully compare files of different pixel count they need to be matched for output size. This can be done by downsizing the larger file or upsizing the smaller one.
I did both using Bicubic Sharper in Photoshop.
Note: there is no useful purpose to be served in comparing different sized files. Of course the larger files will appear more grainy but that is not a fair or realistic comparison. At some point in the history of any picture it will be output at a particular size.
I also made minor adjustments to color balance and lightness so the files were as accurately equivalent as I could make them.
Results-Resolution
I found the TZ90 had a slight advantage in resolution when either the FZ300 files were upsized or the TZ90 files downsized.
Note that I had to closely inspect the files side by side at 100-200% on the monitor to confirm this.
In the real world of photography many factors completely unrelated to pixel count will determine whether a photo is perceived as having good detail resolution or not.
These include camera shake, subject movement, lighting, type of subject, output picture size, exposure, light levels…..etcetera…..
I rate pixel count one of the least relevant factors in determining image resolution or perceived sharpness which is a related but slightly different thing.
Results-Luminance noise
As indicated above noise reduction technology has advanced to the point that color noise can be readily removed from most files even at high ISO settings.
Which leaves luminance noise appearing as grain in the photo.
Higher ISO settings produce more grain.
When I downsized the TZ90 files to match those of the FZ300 in output size I found a slight advantage to the FZ300 at high ISO settings.
When I upsized the FZ300 files to match those of the TZ90 I found there was a slightly different quality to the grain pattern at ISO 3200. The FZ300 files had fewer, larger grains. The TZ90 files had more, smaller grains.
I could not say one was preferable to the other.
Conclusion
The 20Mpx TZ90 delivered a very slight (but also very lens quality dependent) increase in resolution over the FZ300. This would only be evident in a controlled test setting. Resolution/sharpness in the real world is much more dependent on other factors.
The 12Mpx FZ300 produced slightly less luminance noise at high ISO settings but only with one size matching protocol.
Comment
In view of these findings one must wonder why pixel counts continue to rise.
I can only imagine the reason is marketing.
There are still plenty of review sites which list the pixel count of cameras with the suggestion that more is better.
With respect to the two cameras compared in this post it would appear that unless photographers spend their lives photographing test charts more pixels are not better but not worse either.