Re: Triplet after Trigger

["M.Stein" <matthias@gom.com>]

Hello Christian!

Sorry for the late answer, but it's sometimes difficult to switch if
working on comletely other things (like microcontrollers urggg...).

On Thu, 11 Jan 2001, Christian Schulz (ahlrich) wrote:

> Hello Mr. Stein!
> 
> 
> I managed to get the ouput of an failed snap.
> And I made some tests, of what can be wrong.
> I found out, that a failed snap looks like this:
> 
> 	picture 1)  picture 1 or 2 of this tripplet
> 	picture 2)  picture 3 of this tripplet
> 	picture 3)  picture 1 or 2 of NEXT tripplet
> 
> I can't decide if picture 1 or 2 of one tripplet, because picture 1 and 2
> of one tripplet are the same, only picture 3 is different.
> 
> I think, the only difference in the output of a snap that were Ok and a
> failed snap are the two following lines:
> 
> 	Jan 11 14:09:12 faui5pc22 kernel: icpci[W]: sample current field in VB?
> 	Jan 11 14:09:12 faui5pc22 kernel:
> 
> But I don't know, what they mean.

The following explanation is based on the assumption, that you are using
an INTERLACED camera. If not, we have to think again about the problem.

If you are trigger an interlaced camera, it can happen while grabbing an
odd OR even field. But the starting field of the grabber defaults only
to the even one. So if you want to be fast, you have to switch the
starting field to odd, if you are into an even field.
But there is a problem - you have to know in which field you are. The
field signal is stripped from the composite signal and changes usually
during the VSync interval. So I have added the warning (you have seen),
when sampling the field signal during a possible switching of this signal
to avoid race conditions between hard- and software. (The hardware thinks,
we are into a odd field, while the software thinks it is even).
For a first test you can comment out the lines 340 to 366 in
icpciMain.c (from icp_splhigh () to icp_spllow () including this two
statements).  
 
						matthias

> I attached the whole message output.
> 
> 
> I hope, you can find the error.
> Best regards
> 
> Christian
> 
-------------------------------------------------------------------------------
GOM mbH                                                 Tel.: ++49 531/3804-330
Gesellschaft fuer optische Messtechnik                  Fax.: ++49 531/3804-333
Rebenring 33                                            Mail: M.Stein@gom.com
38106 Braunschweig GERMANY                              http://www.gom.com
-------------------------------------------------------------------------------