Image download time

Been having this problem for a while now, I use a QSI583WSG and it takes forever to download, approx. 23-30 secs, if I use neb3 it takes seconds, is there any particular reason for this? just makes the autofocus take forever.

Are you binning your AF shots? When using Nebulosity are you binning those exposures?

2x2 in sgp, none in neb.

Well shucks. I was hoping you’d say the opposite. I don’t know what would be causing this.

This is because SGP uses the “normal” speed download for auto focus frames rather than using the high speed method which results in more noise and poor results for auto focus.

We tested both read methods and found that the high speed mode negatively affected the auto focus result. Thus we use the normal speed mode.

If it’s too slow you can up the binning to 3x3 which should help some.

Thanks,
Jared

Thanks Jared, will give it a go next time out.

Isn’t the QSI583 notorious for being a very slow camera? Similar to the original ST8300?

Your download times are what they are.

Thanks Mads0100 i am aware of that but in neb3 it is only a couple of seconds, only in sgp its about 23.

1 Like

Wow. Thanks! I would not have expected it to be that dramatic.

Yes it is as its USB 1.1. My stl-11000 is also extremely slow.

The reason that you see this is because Neb uses the “fast” download mode for frame and focus. SGP will use the fast download speed for frame and focus as well, but for auto focus frames it uses the normal readout to reduce noise.

This is not something that we’re likely to change. Unfortunately the other option is poor focus…

Thanks,
Jared

1 Like

Thanks Jared, I guess I didn’t realize that ‘fast download’ could make such a dramatic reduction in time. I considered buying a QSI583 and decided not to based on it’s download times.

I concur with this… it was my primary reason for upgrade to the 683 model.

1 Like

Think if i was buying new then would be good reason to but had mine at a good price used so can’t complain, other than that it does its job and thats all that matters in the end.

Tony,

Agreed. It’s a fantastic chip/camera minus that one sticking point.

I am seeing a similar issue with SGP and my Starlight Xpress Trius SX694 camera. Tonight on average it is taking 1 minute from the time a single 600sec 1x1 exposure finishes until it is downloaded in SGP and dithering event kicks off. This entire time it just says “Downloading”. Attached is a log file.sg_logfile_20150120190927.txt (605.8 KB)

Phil,

Your logs are not showing this… I looked at all your image downloads from start to finish and I am seeing about a 2 second download time. Frame 20 is an example, but they are all like this:

[1/20/2015 10:26:50 PM] [DEBUG] [Camera Thread] SGM_CAMERA_CAPTURE message received...
[1/20/2015 10:36:56 PM] [DEBUG] [Sequence Thread] Image reported as complete.  Continuing...
[1/20/2015 10:36:58 PM] [DEBUG] [Sequence Thread] =========== Save file took 2183 ms

So your 10 min exposure took 10 min and 6 seconds (integration + download) and then the save finished 2 seconds later. I think what you are seeing is difficulty recovering from your dither (not sure why the status does not update). It is of note that your sequence started on frame 20 here:

[1/20/2015 10:26:07 PM] [DEBUG] [Sequence Thread] ------------- Starting capture frame for event[0] -------------
[1/20/2015 10:26:07 PM] [DEBUG] [Sequence Thread] Created base name for frame capture: Lum_600sec_-23.1C_1x1_2015-01-20_222607_frame20

So your guider took a full 43 seconds to settle to the specification you set. This common for all of your images.

What OS are you using?

You might want to consider coming down on the dither a bit. You peg PHD at 2.5 pixels every time and your mount needs about 45 seconds to recover from this.

Thanks for the suggestions Ken but it isn’t just the dither time. The total time I am observing for image download and dither is around 1 minute 30 seconds to 1 minute 40 seconds, 1 minute of which is just the time while downloading the image. This is confirmed by looking at my PHD2 graph. After the “Downloading” phase is complete SGP shows it sending the Dither command and I see this reflected in my PHD2 graph. There is roughly 1 minute of time passing from when the exposure finishes until I see the image displayed within SGP. I am using Windows 7.

Here is a video to demonstrate what I am seeing. You will notice once the exposure finishes the “Time Elapsed” in the sequence window is at 05:35:14 and the image is displayed and SGP moves on to the Dither at 05:36:14.

https://vid.me/Ynyw

It is worth noting that shooting short exposures don’t seem to take as long. I am doing flats right now and they download in a couple of seconds.

After some more testing the Downloading phase seems to take longer when PHD2 is running and guiding/capturing images. My QHY5 guide camera is connected to the 3 port hub on the back of my SX694 CCD. I thought the hub might be cause but I just tried connecting a flash drive through and was able to push more than enough bandwidth through it. When I just had PHD2 taking 2 second exposures tonight and not guiding a 600sec exposure took over 20 seconds to download. Without PHD2 running and capturing it was 10secs or less every time. Other applications do not seem to exhibit this same issue while PHD is running but I am going to test some more to 100% rule that out.

I traced this down to being an issue when connecting to my laptop’s USB ports. It must be the built in controller or something. The laptop also has an express card slot so I purchased a USB 3.0 express card off of Amazon for $13. I just tested it and when connected to a built-in USB port I got a download time of 24.5 seconds. When I connected it to one of the express card ports I got a download time of 8.5 secs. Everything else was kept the same I just changed ports. It also removed a weird diagonal hot pixel pattern I was seeing in my bias and dark frames. On the built-in port my max adu for a single bias frame was 8183. On the express card port it was 1101.