dither scale setting

[steve]

phd2.6.1 DEV 7 + APT
Hi
I'm using PHD2 to dither a Canon 700D on a 760mm focal length telescope
Guiding is via a st80 and zwo120

In PHD2 I have:
brain > global > dither settings
The default is set at Scale 1.0
(options 0.01 - 100)

In APT I have
Gear > Guide [D] > Guiding Settings
The default is set at Dithering Distance : 1
(options 1 through 5, integer values)

Question, what's a good starting point for these values for my camera?
TIA and clear skies,
Steve

[Andy Galasso]

Hi Steve,

Since you are imaging at 1.17 arc-sec per camera pixel, and guiding at 1.93 arc-sec per guider pixel (calculator), a 1-pixel dither on the guide chip would correspond to about 1.6 pixels on the imaging chip (1.93 / 1.17).   The dither sizes 1 though 5 in APT correspond to 0.5, 1, 2, 3, and 5 pixels in PHD2, multiplied by the dither scale setting. 

For each APT value, here is how much the dither size would be on the imaging chip (assuming you keep dither scale = 1.0):

APT    ZWO     Canon

Value  Pixels  Pixels

1      0.5     0.8

2      1.0     1.6

3      2.0     3.3

4      3.0     4.9

5      5.0     8.2

I would probably choose at least "2" or "3" in APT, but I believe some DSLR users like to use much larger dither sizes to reduce pattern noise (I do not image with a DSLR and don't know much about it, just repeating what I have heard.)

Andy

[steve]

Hi Andy
Thanks so much. That's perfect.
Just one more setting: the Dither mode. I have a choice of random or spiral. Any recommendations?
Thanks again,
Steve

Andy

--
You received this message because you are subscribed to the Google Groups "Open PHD Guiding" group.
To unsubscribe from this group and stop receiving emails from it, send an email to open-phd-guidi...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Andy Galasso]

Random was the traditional dither algorithm in PHD1 and PHD2.  Each dither moves a random distance (up to the dither size) on each axis, in a random direction.

Spiral Dither is newer and moves in a spiral pattern, with each dither being a fixed distance (the dither amount).  There are fewer Dec reversals with spiral dither, so it may be a good choice for a mount with dec backlash.

Andy

[steve]

El 15/07/2016 a las 00:47, Andy Galasso escribió:

Hi again
I just got a note from the guys on the DSS list who recommend 5 to 10 pixels for the Canon. So, say:
 4 in APT and 1.3 in PHD (4.9 x 1.3 = 6.37 pixels) or...
 5 in APT and 1.0 in PHD (8.2 x 1.0 = 8.2 pixels)

Or should I be looking at the ZWO rather than the Canon pixels?

Yes, I know. Sorry.
TIA,
Steve

[Andy Galasso]

Yes, those settings look right to get you in the 5-10 pixel range on the Canon chip.

[steve]

El 15/07/2016 a las 19:19, Andy Galasso escribió:

Yes, those settings look right to get you in the 5-10 pixel range on the Canon chip.

Hi Andy, everyone
Attached is the log of the with APT dither set at 5 and PHD dither set at 1 and random.

Observations:
- The spikes after the dither.
- The dither maximum seems to be 5. I'd sort of expected 8-ish:

APT    ZWO     Canon

Value  Pixels  Pixels

1      0.5     0.8

2      1.0     1.6

3      2.0     3.3

4      3.0     4.9

5      5.0     8.2

Questions:

-Who tells who when to start the next exposure? I'd feel happier if we could wait longer to settle after the dither.
-I'd like to be able to compare the guiding with and without dither but I see that it includes the spikes in it's RMS values.
-Are the spikes at the point of dither normal?

Anyway, the noise is much improved over using darks. Even with the moon last night, the snaps of M13 I took to test the dither are fine. Thanks for getting me this far.
Clear skies,
Steve

[Andy Galasso]

On Sun, Jul 17, 2016 at 5:15 AM, steve <stev...@gmail.com> wrote:

-Who tells who when to start the next exposure? I'd feel happier if we could wait longer to settle after the dither.

I'm pretty sure APT uses the old PHD1-style interface.  APT tells PHD to dither, then polls PHD asking for the current guide error. When the error goes below some threshold APT starts the next exposure.  You should be able to find a setting for the threshold in APT.

 

-I'd like to be able to compare the guiding with and without dither but I see that it includes the spikes in it's RMS values.
-Are the spikes at the point of dither normal?

The spikes *are* the dithers 😀

PHD2 Log Viewer can exclude the dithers and the settling frames,  Open Settings (menu File => Settings) and check "Exclude settling parametrically".  Enter the values (pixels, seconds) to match the settings you have in APT, click OK.

Then right-click the graph and select "Exclude frames settling".

Andy