Microfading: Questioning the Basics

by Andrew Lerwill, IPI Research Scientist

Building upon his PhD work in micro-fading spectrometry, Andrew’s research interests have focused on the use of diverse technologies to measure, predict, and control photochemical damage to cultural heritage, a subject on which he publishes and consults. Prior to joining IPI in 2013, Andrew worked at The Getty Conservation Institute and the Tate Gallery in the UK.

The technique of microfading testing was first introduced to conservation in 1999 by Whitmore et al. (1, 2). Around the same time Pretzel (3) was independently developing a similar micro-spot fading test. Both these instruments can conduct accelerated light ageing on the surface of a real object. As a result they are able to assess fading rates of fugitive colors and help shape display policy. During the last fifteen years, the technique has been tested and studied by conservation scientists and also adapted for new applications and research purposes(4).

A microfadometer functions by measuring induced color change. This happens through direct high intensity illumination of the material on a sub-millimeter scale where spectral measurements are carried out simultaneously. Within the small illuminated area, fading continues only to a certain level that is not discernible, making it possible to directly test an object’s light sensitivity. As such it has been used by major cultural heritage institutions worldwide to directly assess the light sensitivity of their collections.

At the Tate Gallery we adapted the technique for experimental purposes in an investigation of the effect of light exposure on pigments in low-oxygen environments (in the range 0-5% oxygen)(5). To do this a purpose-built automated microfadometer was constructed for testing hundreds of samples, including multiple samples of traditional watercolour pigments from 19th and 20th-century sources selected for concerns over their stability in anoxia.

During my Post-Doc I followed ten years of collaborative work between conservation scientists at the Getty Conservation Institute and conservators from the Getty Research Institute and the J.P Getty Museum in testing the light sensitivity of collection items with a microfading tester. During this time it was decided to start transferring testing activity from the scientists to the conservators. This involved transfer of knowledge and expertise in order to run the test and interpret the resulting data. My contribution was to design a new fadometer to be used as a simple light sensitivity screening tool for conservators. We designed and tested a portable microfading tester (named µ-MFT) devised for maximum simplicity of engineering, small size, reduced cost, and minimal user training. µ-MFT is in use there today (6) and can be seen in the image and design schematic below.

Based on the last eight years of my research in microfading I think it is safe to say that a wide diversity of opinion exists on the technique. I have found that an issue one individual considers of negligible importance is a sufficient justification for raising serious concerns for another, or occasionally even for dismissing the technique completely. This makes developing (or even discussing) the technique difficult and inhibits progress. Conducting experiments that address and answer concerns about the microfading technique becomes more complex due to a lack of consensus regarding specific problems or areas of concern. The diversity of viewpoints typically stem from the various backgrounds of the people in the field concerned with this issue. As a result, development of a comprehensive, broadly accepted experimental appraisal is a challenge.

To determine if the technique works we must first ask what “work” means to the user. The intended use has often diverged from the initial intent of the technique. The qualitative and quantitative outcomes that result from microfading measurements can be used in a variety of ways:

  • We can question if the object under test is light sensitive.
  • We can apply the technique to differentiate the degree of light sensitivity relative to that of ISO Blue Wool Standards (7).
  • We can aim to accurately predict quantitative future color change in low luminance conditions.

Each of these outcomes is increasingly complex to achieve. All accelerated tests by their nature involve extrapolation which ideally requires justification from detailed physical or chemical knowledge of the effect of the accelerated variable on the degradation mechanism. This knowledge is rarely available or cannot be simply or practically determined (8).  When consulting the literature, approximately 2/3rds of the materials can undergo accelerated photochemical degradation, with results for the same light exposure independent of time of exposure or the intensity of light used (9) (known as the reciprocity principle (10)).

Concerns associated with microfading techniques may include potential biphotonic events, diffusion-limited photo-oxidation reaction rates, dehydration and heating of the sample, the choice of different color difference units, the variability of using and measuring ISO Blue Wool Standards, variation in spectral power distribution between lamps used in object display and the lamp used in accelerated light aging, human and instrumental error in color measurement, the small sampling area, the required number of samples, the length of fading time and the potential interaction of individual chemical components of the object tested.

With all the complexity it could be argued that a simple comparison of microfading results with more accepted methods of accelerated light aging is helpful.  During my time at the GCI a comparison was made between the 2.2Mlux dose from a microfadometer to that of a QUV Weatherometer light aging chamber (with UV filtration).  Ten different Aniline dyes supplied by Dr. PH Martin (2 drops of dye with 5ml distilled water on Whatman filter paper) were each faded for 10 minutes using the microfadometer and then for 21 hours using a QUV Weatherometer (with UV filtration) so the samples received the same Lux-hours dose. See the table below for a comparison of results shown in DE2000 color difference units.

(product code)

(Light box)


Scarlett (5a)



Juniper Green (12a)



Moss Green (24b)



Persimmon (3a)



Wild Rose (19b)



Amber Yellow (16b)



Daffodil Yellow (15b)



Turquoise Blue (8a)



Grass Green (11a)



Moss Rose (7a)




If we plot the results of microfading on a y axis, and lightbox aging on the x axis, we see a correlation. A relationship exists between the two fading regimes (as shown by the solid line) but the same light dose did not result in a similar result in terms of color difference (results would follow the dashed line if this were the case).

The order/ranking of results was very similar in both cases and the ability of the microfadometer system to rank the dyes in terms of light fastness was shown to perform well (with the exception of Moss Green (24b) and Juniper Green (12a) which were out of order.


Order/ranking of results






















Light box ageing












From this experiment it appears that the difference of light intensity and technique used in the two methods of light accelerated aging did not significantly change the ranking between various sensitive materials. The microfadometer testing does not seems to induce a new regime of fading that differs from the one induced by the tests in light aging chambers. However the degree of color change generally appears to be higher when fading occurs under lower luminance conditions using a light box.

Based on these results we can ask “does microfadometer testing work?”

  • We can confidently question if the object under test is light sensitive.
  • We can cautiously apply the technique to differentiate the degree of light sensitivity relative to that of ISO Blue Wool Standards or other colorants in the test.
  • We cannot aim to accurately predict quantitative color change in low luminance conditions from microfadometer exposure but only from comparison to the equivalent known fading behavior of Blue Wools.

A similar experiment comparing microfading results and lux levels close to typical museum lighting conditions (250 lux) yielded similar results and conclusions and will be published.

Predictions of quantitative future color change may be possible with more research, allowing us to empirically determine a common relationship to link microfading with lightbox aging and real low luminance photochemical aging in a display setting.

There are exciting opportunities for further exploration of the accuracy of the technique as well its limitations. The technique is sure to be more widely employed in cultural institutions, by preservation professionals, and in experimental applications.


  1. Whitmore, P.M. Pan, X. Baillie, C. Predicting the fading of objects: Identification of fugitive colorants through direct nondestructive lightfastness measurements. Journal of the American Institute of Conservation 1999, 38, 395-409.
  2. Whitmore, P.M. Baillie, C. and Connors, S.A. Micro-fading tests to predict the result of exhibition: progress and prospects. Tradition and Innovation: Advances in Conservation. A. Roy and P. Smith, eds. International Institute for Conservation: London, 2001, 200-20.
  3. Pretzel, B., Determining the colour fastness of the Bullerswood carpet. A. Roy and P. Smith, eds. Tradition and innovation: advances in conservation. London: IIC, 2000, 150–154.
  4. Ford, B.  Resources - Microfade Services. Available online at: http://microfading.com/resources.html [Accessed 3 June 2014]
  5. Lerwill, A. Townsend. J.H. Thomas, J. Hackney, S. Liang, H. Photochemical colour change for traditional watercolour pigments in low oxygen levels.  Studies in Conservation 2014
  6. Pesme, C. Lerwill, A. Beltran, V. and J. R. Druzik. Development of a contact portable MFT to assist display decision of light sensitive collection items.  Journal of the American Institute of Conservation (in review).
  7. British Standards Institution. Methods of test for colour fastness of textiles and leather, BS 1006: 1978. BSI: London.
  8. Escobar, L.A. Meeker, W.Q. A review of accelerated test models, Statistical Science 2007, 21 552-77.
  9. Martin, J. Chin, J. Nguyen, T. Reciprocity law experiments in polymeric photodegradation: a critical review. Progress in Organic Coatings 2003, 47, 294.
  10. Bunsen, R.W.  Roscoe, H.E., Photochemische untersuchungen, Annalen der Physik und Chemie., 1859, 108, 2, 193.