Limits...
Poggendorff rides again!

Ekroll V, Gilchrist A, Koenderink J, van Doorn A, Wagemans J - Iperception (2015)

Bottom Line: Here, we consider the occlusion of a subjectively linear ramp of tonal values.In a simple experiment, we find results closely resembling those of the geometrical Poggendorff.Yet, the "explanations" offered for the latter hardly apply to the former case.

View Article: PubMed Central - PubMed

Affiliation: Laboratory of Experimental Psychology, University of Leuven (KU Leuven), Leuven, Belgium, e-mail: vebjorn.ekroll@ppw.kuleuven.be.

ABSTRACT
The Poggendorff illusion is one of the most exhaustively studied illusions. Can it be revived as an interesting problem? Perhaps by moving it to a slightly different domain. Here, we consider the occlusion of a subjectively linear ramp of tonal values. In a simple experiment, we find results closely resembling those of the geometrical Poggendorff. Yet, the "explanations" offered for the latter hardly apply to the former case. Depending upon one's perspective, this may be taken to "revive" the Poggendorff illusion.

No MeSH data available.


Related in: MedlinePlus

At left: The average of the three sessions by four of the authors (author JW had a qualitatively different effect, although this would hardly show up in the average). We used bilateral adjustment. The result looks much like a regular “Poggendorff illusion.” The Y-axis represents the 256 gray levels on the computer monitor rescaled to (0–1). The red curve (occluded part dashed) shows a grayscale ramp which appears as a linear brightness scale when it is not occluded. When occluded in the middle, however, the lower part appears too dark (or, equivalently, the upper part appears too light) for a subjectively linear brightness scale‥ The blue curves show the readjustment of the two visible parts of the brightness ramp necessary to correct for this illusion. At this setting, the two visible parts appear as a single linear brightness scale partially hidden behind the occluder. The curves shown here are the pointwise averages of the brightness ramps at the final settings chosen by the observers in terms of device gray values. At right: The geometrical Poggendorff in the same format, although with a (conventional) unilateral adjustment.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4441018&req=5

Figure 2: At left: The average of the three sessions by four of the authors (author JW had a qualitatively different effect, although this would hardly show up in the average). We used bilateral adjustment. The result looks much like a regular “Poggendorff illusion.” The Y-axis represents the 256 gray levels on the computer monitor rescaled to (0–1). The red curve (occluded part dashed) shows a grayscale ramp which appears as a linear brightness scale when it is not occluded. When occluded in the middle, however, the lower part appears too dark (or, equivalently, the upper part appears too light) for a subjectively linear brightness scale‥ The blue curves show the readjustment of the two visible parts of the brightness ramp necessary to correct for this illusion. At this setting, the two visible parts appear as a single linear brightness scale partially hidden behind the occluder. The curves shown here are the pointwise averages of the brightness ramps at the final settings chosen by the observers in terms of device gray values. At right: The geometrical Poggendorff in the same format, although with a (conventional) unilateral adjustment.

Mentions: The global average reveals a very significant “Poggendorff effect.” However, it is not quite clear-cut: Although four of the authors showed very similar results, one author had the opposite effect. This proved reproducible, thus we have to leave an open end here. In Figure 2, we show the result on omission of this participant, although there is hardly a visible difference with the full result. Anyway, the effect is huge (Michelson contrast 23 ± 13%), and a graph of the result shows great similarity to the original Poggendorff. Note that device intensity values (rather than “linear” brightness f(x)) are plotted on the y-axis of Figure 2 left. Plotted in terms of brightness, the curves would be straight rather than slightly curved.


Poggendorff rides again!

Ekroll V, Gilchrist A, Koenderink J, van Doorn A, Wagemans J - Iperception (2015)

At left: The average of the three sessions by four of the authors (author JW had a qualitatively different effect, although this would hardly show up in the average). We used bilateral adjustment. The result looks much like a regular “Poggendorff illusion.” The Y-axis represents the 256 gray levels on the computer monitor rescaled to (0–1). The red curve (occluded part dashed) shows a grayscale ramp which appears as a linear brightness scale when it is not occluded. When occluded in the middle, however, the lower part appears too dark (or, equivalently, the upper part appears too light) for a subjectively linear brightness scale‥ The blue curves show the readjustment of the two visible parts of the brightness ramp necessary to correct for this illusion. At this setting, the two visible parts appear as a single linear brightness scale partially hidden behind the occluder. The curves shown here are the pointwise averages of the brightness ramps at the final settings chosen by the observers in terms of device gray values. At right: The geometrical Poggendorff in the same format, although with a (conventional) unilateral adjustment.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4441018&req=5

Figure 2: At left: The average of the three sessions by four of the authors (author JW had a qualitatively different effect, although this would hardly show up in the average). We used bilateral adjustment. The result looks much like a regular “Poggendorff illusion.” The Y-axis represents the 256 gray levels on the computer monitor rescaled to (0–1). The red curve (occluded part dashed) shows a grayscale ramp which appears as a linear brightness scale when it is not occluded. When occluded in the middle, however, the lower part appears too dark (or, equivalently, the upper part appears too light) for a subjectively linear brightness scale‥ The blue curves show the readjustment of the two visible parts of the brightness ramp necessary to correct for this illusion. At this setting, the two visible parts appear as a single linear brightness scale partially hidden behind the occluder. The curves shown here are the pointwise averages of the brightness ramps at the final settings chosen by the observers in terms of device gray values. At right: The geometrical Poggendorff in the same format, although with a (conventional) unilateral adjustment.
Mentions: The global average reveals a very significant “Poggendorff effect.” However, it is not quite clear-cut: Although four of the authors showed very similar results, one author had the opposite effect. This proved reproducible, thus we have to leave an open end here. In Figure 2, we show the result on omission of this participant, although there is hardly a visible difference with the full result. Anyway, the effect is huge (Michelson contrast 23 ± 13%), and a graph of the result shows great similarity to the original Poggendorff. Note that device intensity values (rather than “linear” brightness f(x)) are plotted on the y-axis of Figure 2 left. Plotted in terms of brightness, the curves would be straight rather than slightly curved.

Bottom Line: Here, we consider the occlusion of a subjectively linear ramp of tonal values.In a simple experiment, we find results closely resembling those of the geometrical Poggendorff.Yet, the "explanations" offered for the latter hardly apply to the former case.

View Article: PubMed Central - PubMed

Affiliation: Laboratory of Experimental Psychology, University of Leuven (KU Leuven), Leuven, Belgium, e-mail: vebjorn.ekroll@ppw.kuleuven.be.

ABSTRACT
The Poggendorff illusion is one of the most exhaustively studied illusions. Can it be revived as an interesting problem? Perhaps by moving it to a slightly different domain. Here, we consider the occlusion of a subjectively linear ramp of tonal values. In a simple experiment, we find results closely resembling those of the geometrical Poggendorff. Yet, the "explanations" offered for the latter hardly apply to the former case. Depending upon one's perspective, this may be taken to "revive" the Poggendorff illusion.

No MeSH data available.


Related in: MedlinePlus