Re: [M]: Re: SCSI vs Serial speeds
| Subject: | Re: [M]: Re: SCSI vs Serial speeds |
| From: | Chris Heapy |
| Date: | Tue Mar 10 09:22:00 1998 |
Geoff Shapiro wrote:
>
> Chris:
>
> I do not own a CCD camera and have never done CCD photography but I have
> followed these conversations and I have always wondered how much smarts
> the CCD cameras actually have on them.
>
> Do they have an embedded microprocessor? ...or are they just dumb
> devices and controlling software on the PC does all the work? If the
> latter, you gain nothing (actually youd lose) but if the former, Ive
> wondered why the developers of the cpu code dont build in a simple
> non-lossy compression scheme on the data. Perhaps there is enough noise
> in the data to prevent any significant compression, but if there are
> enough identical pixel values this should be a big win for serial
> downloads, although the increase in speed for SCSI downloads may be
> worthwhile, negligible or even negative, depending on the SCSI card and
> computer being used!
>
> Geoff Shapiro
>
> Chris Heapy wrote:
>
> >Oh, another thought: that last comment about the delay due to
> preparing
> >image for download - I noticed that the focussing routine in the
> latest
> >6.4b2 has a very fast refresh time indeed (albeit with a sub-frame), so
> >the delay noted above cannot be intrinsic to the firmware of the
> >camera/controller box. Perhaps there IS a software solution to
> >increasing SCSI download speed.
Well Im groping in the dark here, but certainly the 416 has a
controller box which runs a local code (the 3.1 flash code in my case),
and the box has all the camera control functions within it - though
these appear to be redundant as most if not all the camera settings are
accessed through Pview. I vaguely remember reading about an
improvement in V6.4b of Pview, where it is stated that the focussing
refresh was improved through modifying the Pview code. I can see this
improvement by setting the refresh delay time to zero, and the new
pictures appear at around 1 second intervals - even with quite large
sub-frames. Whether these sub-frames have defects not seen in normal
full-frame transfers I dont know. However, it should in theory be
possible to significantly speed up normal full-frame transfers based on
the speed of SCSI data transport at 5Mbit/sec (min).
The problem is, Meade are not very forthcomming with the specs or
details of interfacing to the Pictor cameras, and what pre-processing
occurs immediately prior to download (Ive heard rumours about
correcting for chip faults but find that difficult to believe).
Impossible therefore to estimate the potential improvements by modifying
Pview alone.
Chris
--
*Model Engineer Support Page*
Engineering: http://easyweb.easynet.co.uk/~chrish/homepage.htm
Astro Stuff: http://easyweb.easynet.co.uk/~chrish/lx200.htm
>
> Chris:
>
> I do not own a CCD camera and have never done CCD photography but I have
> followed these conversations and I have always wondered how much smarts
> the CCD cameras actually have on them.
>
> Do they have an embedded microprocessor? ...or are they just dumb
> devices and controlling software on the PC does all the work? If the
> latter, you gain nothing (actually youd lose) but if the former, Ive
> wondered why the developers of the cpu code dont build in a simple
> non-lossy compression scheme on the data. Perhaps there is enough noise
> in the data to prevent any significant compression, but if there are
> enough identical pixel values this should be a big win for serial
> downloads, although the increase in speed for SCSI downloads may be
> worthwhile, negligible or even negative, depending on the SCSI card and
> computer being used!
>
> Geoff Shapiro
>
> Chris Heapy wrote:
>
> >Oh, another thought: that last comment about the delay due to
> preparing
> >image for download - I noticed that the focussing routine in the
> latest
> >6.4b2 has a very fast refresh time indeed (albeit with a sub-frame), so
> >the delay noted above cannot be intrinsic to the firmware of the
> >camera/controller box. Perhaps there IS a software solution to
> >increasing SCSI download speed.
Well Im groping in the dark here, but certainly the 416 has a
controller box which runs a local code (the 3.1 flash code in my case),
and the box has all the camera control functions within it - though
these appear to be redundant as most if not all the camera settings are
accessed through Pview. I vaguely remember reading about an
improvement in V6.4b of Pview, where it is stated that the focussing
refresh was improved through modifying the Pview code. I can see this
improvement by setting the refresh delay time to zero, and the new
pictures appear at around 1 second intervals - even with quite large
sub-frames. Whether these sub-frames have defects not seen in normal
full-frame transfers I dont know. However, it should in theory be
possible to significantly speed up normal full-frame transfers based on
the speed of SCSI data transport at 5Mbit/sec (min).
The problem is, Meade are not very forthcomming with the specs or
details of interfacing to the Pictor cameras, and what pre-processing
occurs immediately prior to download (Ive heard rumours about
correcting for chip faults but find that difficult to believe).
Impossible therefore to estimate the potential improvements by modifying
Pview alone.
Chris
--
*Model Engineer Support Page*
Engineering: http://easyweb.easynet.co.uk/~chrish/homepage.htm
Astro Stuff: http://easyweb.easynet.co.uk/~chrish/lx200.htm
