Some precise details:
If the horizon were badly estimated these results can be shifted one day or two. The same if one goes up in altitude, for example, if one observes since the first stage of the castle, the results can also shift one day. It would be advisable to take photographs to check these forecasts. Moreover these calculations were made with a certain value of the horizontal atmospheric refraction, but the refraction varies with the atmospheric conditions (pressure, temperature...) the moments of sleeping can thus vary slightly compared to the predicted values. We gave these values to the second of time near but it is only the internal precision of calculation, the precision of the prediction counts held of uncertainties on the refraction is of about a minute of time.

If one studies the data above one notes that the azimuth of the center of the Sun to its sleeping never falls exactly into the axis from the channel, it is nearest for the median dates (on April 27 and on August 15), it is slightly in the south (towards the left) when the azimuth is lower than 111°49' 56 "and slightly in north (towards the line) when the azimuth is higher than 111°49' 56". In all the cases it is possible to take a photograph in the axis of the large channel but the center of the Sun will be more or less high compared to the horizon. One can also exploit the height of the observer while going down or while rising compared to the terrace of the frontage. If one goes up, the center of the Sun will be higher, if one goes down, the center of the Sun will be lower.

Caution:
If the Sun with its sleeping dazzles you, do not look at it directly, it is that it is still too high on the horizon. In this case prevent photographing it without filter, you are likely to damage your camera and your sight if you use an apparatus with aiming reflex.