Certain Truths About Amateur Film

How poor is small gauge film –Really!!

The fact is that that old 8mm and Super 8 mm film (and yes even some 16mm) is not all that it is cracked up to be in terms of its supposedly inherent super properties.  There is lots of talk in the forums, other discussion groups and blogs about how 8mm film has great resolution and high dynamic range. I say phooey!!

Yes there is film stock like Fuji’s Velvia line and Kodak EKTAR 100 that exhibits great colour saturation, exceptional contrast and of course fine grain properties, but as for the rest (if not bulk) of the consumer grade film stock, it really begs the question, where did these notions come from?  Eight mm film even in its untarnished state does not exhibit high resolution and does not have a high dynamic range as one is lead to believe.  I touched on this on another post in this blog.

Lets break this down.  The resolution of this size of film gauge (even though the same film stock could be used in the higher gauges like 16mm and above), is really measured by its finest level of grain.  When measured in terms of line pairs per millimeter, the standard metric, 8mm at best is claimed to be better than 80lpm to 100lpm, for a resolution of about 700 to around 900 horizontal lines. When captured using a DV AVI system in standard definition, the maximum line resolution is 550 lines for DV.  So yes there will be image decimation. An HD system in 1080p will do more than capture the finest details in this case.

ref:  8mm film frame =  4.5mm x 3.3mm;     Super 8 film frame  = 5.8mm x 4mm

In practice however and in many cases, the line pair count is inherently in the 30-50lpm range despite the poor quality of the film itself adding to the deficit.  Films in this category will just meet the sampling range of SD DV – just good enough.  Will HD digitizing do a better job, sure, but why?  The film quality is poor and the dynamic range is low, which gets me to the next overblown property of small gauge film, dynamic range.

What is dynamic range?  Not to be confused with contrast ratio which is the ratio of the darkest to the brightest elements in an image, all without crushing and clipping respectively quoted in db.  Dynamic range is more subtle.  It is the measure of how well an image sensor can discriminate  shades in an image being digitized at the smallest quantized bit level – from the noise floor to maximum white.  Higher ratios suggest the capture and reconstruction of an image with all of its finest details intact and maintaining all the shades in between.  High pixel quantization levels can achieve this assuming the image depth within the film frame is there to capture; 8mm doesn’t have it.  Ratios like 60db or more are quoted at times.  A ratio of 60db, says a telecine system can resolve up to 1000 pixel shade differences.

A high dynamic range of 60db can be comfortably attained for a photo using 12 bits of pixel quantization, but not video from 8mm telecine.  Standard SD  and HD video uses 8 bits per pixel, for a maximum of 48db+ or 256 levels of shading (assuming a gamma of 1).  If this full range can be attained the video will look pretty damn good, but not 60db worth. In many cases small gauge film stocks offer mainly flat and washed images, dramatically reducing a high quality experience where even the performance boundaries of an average frame by frame telecine system are not stressed.  There is no discussion yet on the effects of gamma adjustment on dynamic range.  Another blog comment.

Professional telecine systems for 35mm cinematic films can get 12 bit pixel accuracy and be tweeked for 60db performance, but don’t count on it for 8mm films.  Bottom line, in many instances, 8mm film transfer using an SD frame by frame scanning will provide for quite good results and save money to boot.

Having a telecine system with 12bpp capability, incorporating high end mega pixel lenses and 2K-4k resolution imagers for 16mm and larger film gauges coupled with film restoration techniques applied like dustbusting and scratch removal is quite another discussion.

Leave a Reply

Your email address will not be published. Required fields are marked *