For 60 % sidelap and the control distribution H4 (see figure 4) Krack’s results showed the following 
improvements in oz mean and oz max respectively. 
07 mean oz max 
12253 factor 1.1 factor 1.0 
1-7 factor 1.3 factor 1.1 
These preliminary comparisons of theoretical accuracy are based on wide angle photography. 
When using superwide angle photos the bundle block results let expect the same accuracy in plani- 
metry but a much better one in height (the standard deviations of the z coordinates of adjusted minor 
control points decrease with the ratio of the calibrated focal lenghts). This result is in disagreement 
with the results for model blocks where the height was not improved using superwide angle. Wether 
this is realistic or not must be checked by further empirical tests. 
5. Conclusion 
Among the control distributions investigated those seem to be most important which give an 
absolute accuracy after blockadjustment which is approximately in the same order as the single 
model accuracy. To reach that in planimetry the well known perimeter control with a point density 
of 2 to 4 base lengths is sufficient. 
To obtain a comparable height accuracy a dense perimeter control (distance 2 baselengths) is 
necessary and in addition a relatively dense net of height control points within the block (2 baselengths 
across strip direction and approximately 4 baselengths in strip direction). 
The use of 609, sidelap can really be recommended because it improves the accuracy obtainable 
with 20 95 sidelap by a factor 1.4 to 2.0 and in special cases even more. As far as it's possible from the 
stand point of identification the disadvantage of twice the number of models can be eliminated by a 
reduction of the photoscale by a factor 1.4. Then in comparison with 20 9$ sidelap at the original 
scale the accuracy is still better or at least the same and a remerkable number of control points is 
saved due to the larger model base. 
Wide angle and superwide angle photography seems to give approximately the same accuracy 
after the block adjustment starting from the same photoscale. This is valid at least for blocks with 
independent models. Whether in bundle blocks superwide angle gives a better hight accuracy must 
be checked empirically. 
Theoretical accuracy investigations normally assume errorfree control. Of course the accuracy after 
block adjustment changes to the worse when the control points have certain standard deviations. As 
an example has shown fortunately the loss of accuracy is only in the order of a few percent as long as 
the control is not less accurate than photogrammetry. 
Bundle block adjustments let expect significantly better results than block adjustments by inde- 
pendent models. In case of a fairly dense control a change to bundles improves the planimetric 
accuracy more than the height accuracy. 
Zusammenfassung 
Für große räumliche Blöcke mit unabhängigen Modellen bzw. Bündeln wird die theoretische Ge- 
nauigkeit in Abhängigkeit von der Blockgröße und von der Paßpunktanordnung untersucht. Als 
Ergebnis werden Genauigkeitsgesetze für die mittlere Streuung und für die maximale Streuung der 
ausgeglichenen Lage- bzw. Höhenkoordinaten angegeben. 
Resume 
Pour de grands blocs spatials avec respectivement des modéles indépendants et des gerbes per- 
spectives on étudie la précision théorique en fonction de la largeur des blocs et de la disposition des 
points de contrôle. Comme résultat on donne des lois de précision pour la dispersion moyenne et 
pour la dispersion maximale des coordonnées planimétriques et verticales ajustées. 
10