[Elphel-support] Fwd: combining frames

[Andreas Bean]

Hello Oleg!

I got your mail and tested it yesterday.

It's working !!PERFECTLY!!

Thank you very much! The next step which I will do is the geometric
conversion to a panorama-frame on the pc. I will keep you informed.

Thank you Oleg,

Andreas


Oleg Dzhimiev schrieb:
> Hi Andreas,
>
> I'm not sure you got my last letter. Have you tried? Does it work?
>
> Oleg
>
> ---------- Forwarded message ----------
> From: *Oleg Dzhimiev* <andersonnotgood at gmail.com
> <mailto:andersonnotgood at gmail.com>>
> Date: 2009/12/8
> Subject: Re: [Elphel-support] combining frames
> To: Andreas Bean <office at beanbox.com <mailto:office at beanbox.com>>
> Cc: elphel support-list <support-list at support.elphel.com
> <mailto:support-list at support.elphel.com>>
>
>
> Hi Andreas,
>
> You were getting bad frames because in my scripts I supposed you set
> the resolution like on the picture attached and you did it by scaling
> - the camvc writes different values to sensors in these 2 cases and
> I'm using these values to calculate the combined frame parts - my fault.
>
> So, could you set the resolution like shown on the picture and not in
> tab 5?
> I also put an input box for scaling to 10359_controls - 2 for 1/2, 3
> for 1/3 and so on.
> Tested my script - it works for scale factor 2 (I need to test other
> factors) and also for 1 at different resolutions. Sending you files
> and x359.bit - the one you got from the CVS.
>
> The procedure:
> 1. set phases
> 2. set trigger
> 3. set the correct result frame resoluton in camvc
> 4. in 10359_controls - set scale factor to 2, check combined box if
> it's not checked and start.
>
> Do you need binning/averaging also?
>
> Oleg
>
> 2009/12/8 Andreas Bean <office at beanbox.com <mailto:office at beanbox.com>>
>
>     Hello Oleg,
>
>     I just tested the new mode with the sources from cvs. I updated
>     the files:
>
>     10359_controls.html
>     10359_modes.php
>     test_inner_10359.php
>     x359.bit
>
>     ------------ x359.bit v1.9 from cvs ---------------
>
>     After the following procedure
>
>     1) camvc, set on tab 5 horizontal to half, vertical to full ->
>     1296x1920
>     2) set cam to triggered mode using
>     http://10.0.0.9/parsedit.php?embed=0.1&title=External+trigger+controls&TRIG=4&TRIG_PERIOD=19200000&TRIG_DELAY&EARLY_TIMESTAMP=1&THIS_FRAME=@&TRIG_CONDITION=0&TRIG_OUT=0&SENSOR_REGS30=@
>     <http://10.0.0.9/parsedit.php?embed=0.1&title=External+trigger+controls&TRIG=4&TRIG_PERIOD=19200000&TRIG_DELAY&EARLY_TIMESTAMP=1&THIS_FRAME=@&TRIG_CONDITION=0&TRIG_OUT=0&SENSOR_REGS30=@>
>
>     ** result 1 **
>     3) enable combining by pressing the button "Alternation mode with
>     buffering" and having "Combine into one frame" checked
>     http://10.0.0.9/359/10359_controls.html
>
>     --> result: buffered frame is scrambled, buffered and unbuffered frame
>     flipping
>
>     ** result 2 **
>     3) enable combining by pressing the button "Alternation mode"
>     --> result: no combined frame, I guess thats ok
>
>     ** result 3 **
>     3) enable combining by using your previous script
>     http://10.0.0.9/359/10359_combining.php?x=1296&y=960&mode=on
>     <http://10.0.0.9/359/10359_combining.php?x=1296&y=960&mode=on>
>     <http://192.168.0.9/359/10359_combining.php?x=1296&y=960&mode=on
>     <http://192.168.0.9/359/10359_combining.php?x=1296&y=960&mode=on>>
>
>     --> result: position and brightness of the combined frame is unstable
>
>     ---------- x359.bit from your email 20091201 ------------
>
>     combining frames works fine for buffering A + B.
>
>     ================================
>
>     conclusion:
>
>     with the x359.bit from cvs I got the same problems like the ones I got
>     when I testet the "alternation mode with buffering" without combining.
>     Additionally the frame of sensor one and of sensor two are
>     interchanging. Contrary, the version with buffering of A+B worked
>     out of
>     the box. I had no problem at all with it. Unfortunately buffering
>     A+B is
>     not sufficient due to the limited bandwidth.
>     The question is: Does it make sense to keep the version which
>     buffers B
>     only?
>     I propose to move on to A+B/C+D version which improves stability and
>     bandwidth.
>
>     Andreas
>
>     _______________________________________________
>     Support-list mailing list
>     Support-list at support.elphel.com
>     <mailto:Support-list at support.elphel.com>
>     http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com
>
>
>
>
> ------------------------------------------------------------------------
>
> /10359 board controls/
> Use sensors (J2,J3,J4) 	
> Switch to channel : 	
> Sensor test mode ON/OFF 	
>  
>
> Number of phase steps:  
> Channel 0 phase
> Channel 1 phase
> Channel 2 phase
>  
>
> SDRAM clock   	
> Clock phase     	
>  
>
> Combine into one frame 	
> Scale factor 	
> Alternation mode with buffering: 	
>  
> Alternation mode: 	
>  
>  
>
> Request line
>  
>
> 	
> Firmware Version: 	
> Current Channel: 	
> Reg 0x02: 	
> Reg 0x03h(pclk): 	
> Reg 0x03l(sclk0): 	
> Reg 0x05: 	
> Reg 0x06: 	
> Reg 0x07: 	
> Reg 0x20: 	
> Reg 0x44: 	
>
>  
>
>
> ------------------------------------------------------------------------
>

-------------- next part --------------
A non-text attachment was scrubbed...
Name: office.vcf
Type: text/x-vcard
Size: 233 bytes
Desc: not available
URL: <http://support.elphel.com/pipermail/support-list_support.elphel.com/attachments/20091213/8c488fa8/attachment.vcf>