Jump to content

NLCbanner2024.jpg.2478be509670e60c2d6efd04834b8b47.jpg

dropped frames

pipsquip

By pipsquip
in Getting Started With Imaging

Recommended Posts

  • Replies 29
  • Created
  • Last Reply
JamesF

JamesF

Posted
Posted

Some idea of the frame rate and image size that you're trying to use would be helpful.  Also, are you using a relatively new machine for the capture, or something quite old?

James

Link to comment
Share on other sites
pipsquip

pipsquip

Posted
  • Author
Posted

I am using sharpcap 2 I am imaging Jupiter the frame rate is 15 fps and the exposure is about 8 I tried a higher exposure but I didn't get any detail I have tried the capture software that I got with the imager but that was worse

Link to comment
Share on other sites
JamesF

JamesF

Posted
Posted

In sharpcap exposure=8 means 8 seconds I think. Try dropping it to 0.01 (10mS) and increasing the gain.

I think that's very driver/camera dependent.  Some drivers don't offer an exposure control that has any documented relationship with the exposure time.

James

Link to comment
Share on other sites
jambouk

jambouk

Posted
Posted

The book i'm reading (introduction to webcam astrophotography) talks about dropped frames and suggests it happens when the usb cable (in effect) is the bottle neck for all the data trying to get from the camera to the computer. A full hard disc would presumably be a similar bottle neck as it hunted for space to drop them (especially with an old computer).

If your camera allows you to alter the exposure settings and frame rate, try to image at 1/30th - 1/60th of a second (depending on the seeing and your aperture) but image at 5-10 frames per second and see if that reduces dropped frames but improves the quality / detail of each frame and stacked images. Very much trial and error.

If you can reduce the size of the bit of the sensor you are using too, you might be able to get faster frame rates, but as jamesf says, all this depends on the camera/driver age and versions.

The odd dropped frame isn't a problem, and if you get none, it might be a sign you are not pushing the kit as hard as you might be able to.

Good luck

Jd

Sent from my iPhone using Tapatalk

Link to comment
Share on other sites
jambouk

jambouk

Posted
Posted

Jamesf... If they can't alter the exposure time, and they were to select 30 frames per second, presumabely the camera will do an exposure of about 1/30th of a second, and there will be say 20 out of 30 frames dropped, is this a problem as the 10 fos they are getting are shot at 1/30th of a second, or does the system try to start compressing the data and start losing detail?

Jd

Sent from my iPhone using Tapatalk

Link to comment
Share on other sites
JamesF

JamesF

Posted
Posted

Jamesf... If they can't alter the exposure time, and they were to select 30 frames per second, presumabely the camera will do an exposure of about 1/30th of a second, and there will be say 20 out of 30 frames dropped, is this a problem as the 10 fos they are getting are shot at 1/30th of a second, or does the system try to start compressing the data and start losing detail?

What happens is probably entirely dependent on the camera and driver.  I don't think it's possible to say in the general case.  I've no idea what the firmware in the SSIV does.

The SSIV is however basically a QHY5, so perhaps it has similar firmware.  I don't think the QHY5 has any actual control of frame rate in the camera itself.  That's all handled on the PC side.

James

Link to comment
Share on other sites
jambouk

jambouk

Posted
Posted

James, i've just re-read the pages on dropped frames in that book, and the author suggests if lots of frames are being dropped because the the camera runs on usb 1.1..... "But in reality frame rates higher than five per second create so much data that in order to pass it through a usb 1.1 connection the camera is forced to compressed it..."

So is there a way to see how much compression has occurred in the final avi file? Is this shown in the properties of the file, or an advanced property of the file? If so, it might be possible to do some experimenting to see at what frame rate abnormal compression starts to occur and then stick to values below that for any given camera, driver, usb lead, computer combo.

Jd

Sent from my iPhone using Tapatalk

Link to comment
Share on other sites
jambouk

jambouk

Posted
Posted

Ok. It's all out of my comfort zone, but i like experimenting. 10fps just seems a but arbitrary; if 10 really is the cut off where compression starts, then for some systems with old cables, slow hard drives, or whatever, then 8fps might be the cut off as other factors become rate limiting, which is why i wonder if there is a way to tell at what point compression kicks in.

Is all data from a web cam compressed, but at certain frame rates this compression is stepped up to try and get more frames down the usb cable? If so, at what "amount" of compression is compression considered too great for planetary imaging.

I know my logitech 4000 can physically work at nearly 30 frames per second in sharpcap, but half the frames are dropped so only 15 frames per second get put into the final avi file. I'll have to try to compare the size of avi files run at 5, 10, 15, 30 fps and see if it is easy to identify when compression licks in.

I'll wait for jamesf.

Jd

Sent from my iPhone using Tapatalk

Link to comment
Share on other sites
jambouk

jambouk

Posted
Posted

A little test I've just done with my Logitech 4000 (dayight and I've tried to standardise the brightness of what it was pointing at).

Left the sharpcap settings the same (640x480, RGB24, 1/100th second exposures, 20 second video runs); run at 5, 10, 15, 20, 25, 30 fps, each one three times. Taken the average (not scientific I know) of each these, and the number of dropped frames each time. The percentage of dropped frames seems similar (if anything less at 30fps (which was a highly reproducible feature)), and the file size divided by the number of frames in each video was similar, which surprised me, as I'd expected to see a fall in this as compression kicked in. So, I'm not a lot wiser after my experiment, but hopefully one of the clever guys will expalin it or pull my scientific method to shreads :)

JD

logitech 4000 frame rate test.pdf

Link to comment
Share on other sites
allcart

allcart

Posted
Posted

 Out of interest, if you play back your avi frame by frame, do these dropped frames show in the avi as black frames.

When I first got my DMK cams I was getting almost a third of my frames being dropped. Thats 300+ out of a 1000.. During playback, these dropped frames showed as empty black segments lasting about 5 frames or so. After analyzing with AS2 these frames were all at the end.

I solved the problem with a small app from TIS called 'processor idle state manager. I think it turns off unwanted programs while capturing or something.

Anyway, it solved my dropped frame problem.

Link to comment
Share on other sites
JamesF

JamesF

Posted
Posted

Certainly with the SPC900 and perhaps with many other genuine "webcams" the firmware attempts to maintain the requested frame rate by dropping/compressing data in the frame.  The SPC900 does this based on the bandwidth available on the USB bus.  Basically it works out the data rate required to transfer frames to the host PC, compares it with the bandwidth it has been allocated on the USB bus and starts compressing frames when it won't fit.  At 640x480 and 5fps if the camera has been assigned enough bandwidth on the bus then the video stream is unaffected.  At 10fps there's slight compression and beyond that it gets quite aggressive.  A camera with a USB2 interface wouldn't have to be anywhere near so aggressive.  Generally you won't tend to see dropped frames this way, though it is possible.

There's also often a second layer of unavoidable compression with a true video camera because it doesn't provide an actual RGB image.  The data is often sent as luminance and chrominance signals (this is what the IYUV, YUY2 labels refer to) and because the eye is more sensitive to detail than colour, some of the chrominance data is dropped.

Dedicated astro cameras don't tend to work the same way.  They usually provide a full frame uncompressed greyscale or raw colour image.  Often they're not set up to maintain a given frame rate either.  The application says what exposure time is required and data is made available once the exposure is completed.  If the data can't be taken from the camera by the time the next frame is ready it will often be lost.  As long as the host PC can keep taking frames as fast as they're produced that's fine, but after that if the PC can't keep up or there isn't enough bandwidth on the bus to take them all you might miss frames altogether or get broken frames.

There are some other gotchas with some cameras too.  If for instance you half both the height and width of the SPC900 frame, that quarters the amount of data needed to be transferred from the camera.  Do the same with a QHY5 (and perhaps  the SSIV) and I believe it will transfer fewer rows of data for the image, but it still transfers the entire row.  So rather than increasing the potential frame rate by a factor of four you're only increasing it by a factor of two.

James

Link to comment
Share on other sites
JamesF

JamesF

Posted
Posted

 Out of interest, if you play back your avi frame by frame, do these dropped frames show in the avi as black frames.

That's probably application-dependent.  In oacapture if I can see that a frame is broken then it doesn't get displayed or written to the output file.  Some applications may not bother to check whether they have a genuine frame back from the camera and just write the data buffer out regardless.

James

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

Go to topic listing

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.