Hello Yuri,
The reason for this is that when movement correction is applied, the output
has 6 extra channels that give the head position as a function of time.
This is enough to drive the size over 2G.
I think that we need to adjust the maximum file size of the systems
so that measurement is automatically chopped to somewhat smaller
sizes.
The current size is 2147483647 - 100 M. Unfortunately there is no
simple variable where one could set the new value, so it must be put
into the start up script of megacq as an option. Please contact after
sales to figure out formalities needed to in order to get the change done.
Matti K.
Yury Shtyrov wrote:
> Just to add - the problem seems to occur only when movecomp is used
> (but one would rather use it, of course, than not...)
>
> yury
>
> 2009/11/19 Yury Shtyrov <[log in to unmask]>:
>
>> Hello all -
>>
>> I have encountered a new problem with MF2.1 - when the input file is
>> on the upper limit (2G), it refuses to run complaining that the output
>> file is too large (2.1G - why?). As a remedy, it suggests to use short
>> output format, which does not work ('too big data values for format
>> short), or splice the file using skip, which is not great at all.
>>
>> I have tried to downsample the output file to shrink it's size (-ds 4
>> or 3), but MF2.1 still refuses to run for the same odd reason. The
>> very same input file is processed fine in MF2.0.
>>
>> Has anyone encountered this and is there a fix?
>>
>> Thanks!
>> yury
>>
>>
>
>
>
>
|