Thursday, 8 October 2009

two films and a digital

Recently I took my Panasonic G1 my film EOS and 2 films with me to the park for some quick test shots. I was after a few things:
  • some exposed negative and some exposed slide to compare with eachother
  • to then send these films to a fellow to compare his Pacific scanner with my Nikon LS-4000
  • and to compare my G1 with what film can give
I thought when I started putting this together that I'd use 200 ISO as my reference as the G1 is reportedly a native 200 ISO sensor. I picked Sensia (a very commonly available Fuji 200 ISO film) as my negative and Sensia as my slide. I was only able to get this in 100 ISO in my dinky-fart-town here in Finland.

In particularly interested to compare 35mm full frame film with a 21mm wide angle lens prime to the Panasonic with a 9-18mm zoom at 11mm. I had previously seen that my 10D was very close to film but the G1 is a latest model camera and at 12 megapixels in a 4/3 format quite at the cutting edge.

Ok ... here's the slide scanned (and nothing done to it).
[for those who are interested, I have my scanner software set to put the image into Bruce RGB and I then assign that profile when it gets into photoshop. I then convert to sRGB for uploads or use screen grabs, my monitor is Spyder calibrated but I don't know if yours is]



[don't forget to click on any of these images to download the full size]
now, lets zoom in to the patch of red leaves in the middle (to 50% of full resolution).


and compare this to the previous examination of the Neg and the G1



Now firstly I think its important to say that my LS-40 scans a full frame of 35mm to about 5570 pixels wide (4000 Dots Per Inch scan of 1.4 inches is 5600 ... right)

This means that the digital capture of the film will be more greatly magnified than the sensor captured information. So in the previous examination I scaled them to be the same - 4000 pixels wide. I guess that I should do the same here to make direct comparisons with the digital more simple, but the reality is that life is not simple, scans of 35mm film will not neatly fit the mould of digital and so we have to make sence of this somehow.

Probably one of the first things you may notice is how similar the digital is to the negative, but how much darker and inky the slide is. This goes back to the fact that slides are designed with projection onto a screen in mind, while negative is designed to make a print on paper. This is not an insignificant difference.

Ok ... so lets get into pixel peeping and look at 100% views of what we got (100% means one pixel in the file maps to one pixel on your screen.


Slide


which is a segment from an original which is 5573 pixels wide, so it looks bigger. It also looks softer as it hasn't been sharpened.

G1



which is a segment from an image which is 4000 pixels wide and has been sharpened carefully.

It seems to me that the slide image is somehow softer than the image from the digital file (and the image from the negative). Certainly the colour rendition of the digital is more pleasing to my eye than that of the Slide.

One could argue that the digital image seems sharper as it is of a lower magnification, but scaling the slide scan down doesn't seem to make much difference. Possible answers to this could be:
  • scanner focus was off
  • the focus of the lens was off

To answer this I can say that I specifically chose the red bushes as my focus point for scanning, and used Nikons focus system to to that.

On the subject of focusing, I can report I focused carefully on the screen as I could, and I focused manually. As well, Depth of Field calculations for a 21mm lens at f 5.6 suggest that assuming I focused on something which was about 6 meters away (the trees were a bit further away) that the front of the focal zone would be about 3 meters before me through to infinity. This sort of rules out focusing error here.

Interestingly I have observed exactly this effect in every compairson I have done of Negative VS Slide film.

Ok, lastly ... I think that its fair to say that comparing non sharpened scans to post processed digital is a little difficult to do. So below I have done the following:
  • scaled the TIFF from the G1 up to be the same size as the image that comes from the LS-4000 scanner.
  • I have colour balanced the scanned slide to be as close in hue to the TIFF as I could
  • I applied some curves to darken down the shadow details in the TIFF (as the shadow detail was better than the slide was)
  • I applied some strong sharpening and local area contrast enhancement on the Slide scan
[note: the G1 file was a single raw image processed using a flow of dcraw to convert to linear TIFF -> Photomatix to apply careful tone mapping -> photoshop to apply curves and colour balance, there was bugger all sharpening done]


I still think that the G1 comes ahead of the slide, both in colour and detail. As well the Slide actually runs out of dynamic range before the G1 does ... although not before negative did ;-)

note the clarity of the grasses and the leaves. There was no wind on the day.

Certainly the scanned 35mm file will be larger (more pixels) than the G1's file, but the G1 file is so clean and copes with a smidge of upscaling no problems. So when it comes to printing and detail capture it would seem that the G1 and the lenses available to it can at least equal that of the 35mm system. Without a doubt (assuming well exposed images to start with) its easier to get a good printable image from the G1 than it is the 35mm film.

... for my money I'll be using the G1 from now on and using the 35mm for black and white only.

Hope that was useful to someone.

Stay tuned for a scanner comparison in a few weeks!

5 comments:

Noons said...

Good stuff, Chris!
Thanks for publishing this.

Anonymous said...

You've got to compare the dynamic range of a color negative film and the g1. The results will be different.

obakesan said...

Dear Anon

yes, that is also a worthwhile comparison. This is implicitly done in this comparison, and can be seen by looking at the images and seeing the differences in shadow details and hilight responce in the images

Anonymous said...

Looks like the color variation is much better captured by the film camera. I can't tell which is a better representation of the actual scene. Although you might be able to obtain the same results after post-processing the G1's shot.

obakesan said...

Dear Anon

yes, I agree that the film looks better, perhaps even more vibrant. I actually prefer the Negative film more.

I have other related articles in this blog about this exact comparison, so I think I will manage them with key words and allow the chapters to be linked together. I don't really like really long posts