Inside this issue
Graduated ND Filter Testing
Colour Casts and Transitions
Amateur Photographer who plays with big cameras and film when in between digital photographs.
Faced with a large array of graduated ND filters from a large range of suppliers (52 filters from 16 suppliers!) I had to come up with a way of looking at the quality of filters without relying on subjective reading. Fortunately, I had a couple of tools at my disposal: Firstly a high-end flatbed scanner commonly used for museum reproduction, a Fuji Lanovia C-550, and a Greta Macbeth i1 Spectrophotometer for which I had purchased Spectrashop, a tool for analysing and comparing the raw data from the sensor. So after getting my filters in a manageable pile, I proceeded to scan each one and record the spectral distribution of the neutral density side of each graduated filter.
The scan was useful to show the colour of the filter and the way the graduated part looked. To ensure the results were colour accurate, I used a professional calibration transparency (Hutch). In the scans below I increased the contrast and saturation to make the casts more obvious. The actual colour in daylight is better represented by the colour patch next to the transmission spectrum. In order to see how smooth (or not) the graduation of the filter was, I also use a software package called ImageJ which allowed me to draw graphs of density along the length of the filter.
Please note that for the cheaper resin filters where the spectrum is 'lumpy' the light source in the scanner interacted oddly with the dyes in the filters making them look worse than they may appear under daylight conditions. However, this is still a weakness in the filter as colours may shift for some light sources and materials in the same way (e.g. sunsets, plants, artificial lights, etc). We will be testing the filters under daylight/sunlight during our outdoor testing. I plan on testing the filters outside with a Colorchecker passport and adding to this article next week.
The following report looks at each of the filters showing the scan, spectra and graph of density across the filter for each. We'll look at them in alphabetical order by brand. I've held off drawing conclusions based on these results for the moment but it's pretty safe to say that the four or five best filters here will give excellent results and you would be hard pressed to find fault with the next few in a photographic image.
Next issue we'll be looking at how some of these colour casts appear when using the filters outdoors and also at how the filters are to use in the field.
The 84.5 filters are thinner than the typical full 100mm filter at an average of 1.5mm thick over the ones we have, which may cause issues in some holders (we'll check compatibility in another article). They are made of resin and have some of the colour accuracy weaknesses that cheaper dye based filters exhibit. The spectral analysis of most resin filters shows 'peaks' of the different dyes from which they are made. Whilst not as bad as the very cheap resin filters, they are at the lower end of the quality scale. Another thing to note is that resin filters allow infrared to pass through almost unattenuated. This is unlikely to make a big difference in most photographic conditions (as far as I am aware).
Benro filters are made of a glass with a nice, consistent 2mm thickness. The colour cast is good but not perfect with a couple tending toward blue/purple apart from one very accurate example (the 0.6 hard). Definitely at the higher end of quality though.