Experimental Physics and
Industrial Control System

"Mark Rivers" <[email protected]> · Wed, 24 Aug 2011 21:24:12 -0500

With large arrays and slow networks you might need to increase the timeout on caget from the default of 1 second, using the -w option.

Mark

________________________________

From: [email protected] on behalf of Pawel Predki
Sent: Wed 8/24/2011 3:25 PM
To: Chen, Xihui
Cc: TechTalk
Subject: Re: Dynamic images in BOY

Hi Xihui,

Yes, I changed that as well. I get the same amount of data as with
caget. That is 54752 for a 1710x1680 'image' and 16960 for 1000x1000.

Could it have anything to do with the fact that it takes some time to
fill the rec->bptr buffer in the read_wf function? I guess not because I
am able to get a nice 40000 element buffer for a 200x200 'image' but I
get LESS than that for 1000x1000.

I don't know if there's another place some kind of limit is set...

Cheers,
Pawel

W dniu 2011-08-24 19:29, Chen, Xihui pisze:
> Hi Pawel,
>
> Did you set the CSS Prefrences->CSS Core->EPICS->max_array_bytes? CSS doesn't use the environment variable EPICS_CA_MAX_ARRAY_BYTES.
>
> Thanks,
> Xihui
>
>
> -----Original Message-----
> From: Pawel Predki [mailto:[email protected]]
> Sent: Wednesday, August 24, 2011 6:07 AM
> To: Kasemir, Kay
> Cc: TechTalk
> Subject: Re: Dynamic images in BOY
>
> Hello Kay,
>
> It's been quite some time since I raised this topic but now I finally
> managed to try your way for 1-channel images. All in all it works fine.
> There is one problem, though. The images we're dealing width are of
> resolution of up to 1710x1680 which is way above the default
> EPICS_CA_MAX_ARRAY_BYTES value. I changed this environmental variable to
> 6 000 000 and everything seemed fine again. There were no more errors.
> However, only 54752 elements are really being sent. I checked it using
> caget and in BOY.
>
> My database entry looks as follows:
>
> record(waveform, "CAMERA1:WAVETEST") {
>       field(DTYP, "imgwave")
>       field(NELM, "2872800")
>       field(FTVL, "SHORT")
> }
>
> By the way, I tried FTVL, "UCHAR" and I got negative values when the
> real values exceeded 127 when writing to the waveform in the IOC.
>
> I wrote soem very simple device support routines that fill the waveform
> with counter-like data (0..255) in the read_wf routine. I'm setting the
> NORD field of the PV to 1710*1680 and it is there afterwards so I know
> the loop works fine.
>
> Do you have any ideas what can be wrong? Is there some other size limit
> that I am not aware of?
>
> Regards,
> Pawel
>
> W dniu 2011-06-06 15:33, Kasemir, Kay pisze:
>>
>>
>>
>> On 6/6/11 08:59 , "Pawel Predki"<[email protected]>   wrote:
>>> Is it possible to do such a thing in BOY? There is an Image widget but I
>>> believe it's for static images only.
>> The "Intensity Graph" widget displays waveforms that are really grayscale
>> images.
>>
>> For example:
>> waveform with 10000 elements, "width" 100 means:
>> Elements 0...99 are the first row of the image,
>> elements 100..199 are the second row of the image and so on.
>> You need to specify the minimum and maximum element value,
>> like 0 and 255, which will then be mapped to black...white
>> or some other color map of your choosing.
>>
>> -Kay
>>
>
>

--

Pawel Predki
Technical University of Lodz
Department of Microelectronics and Computer Science
ul. Wolczanska 221/223, 90-924 Lodz, Poland
tel. +4842 6312720

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System