Screen refreshing vs integration / internal stacking

[hoops86]

My daily stupid question for you lot

First just to make you’s aware I have no clue about cameras, photography or astrophotography in anyway other than what I have learned here in the VA section.

I cant get my head around this integration (sens up) and internal stacking thing!  

So….. shutter settings…. Lets pick x256 as an example. My simple mind basically assumed this was equivalent to opening the shutter for a certain amount of time. I.e. 2.5 seconds then it would produce an image of 2.5 seconds worth of light. So going up to x512 would be 5 seconds then x1024 is 10 seconds. Is this right?

If that’s right…. That should mean at x256 my image on screen should refresh every 2.5 seconds? At x512 it should display a brand new image of 5 seconds worth of light….

You might have seen me mention in other posts that I just couldn’t push x512 without killing the image. It would give star trails and destroy the clean image I had at x256. My screen would not update every 5 seconds. Its like it would just keep adding to the original image. So a few minutes after jumping up to x512 it was showing me my image from x256 plus all the other x512 frames on top of the x256 frame.

After leaving the camera at x512 for like 30 mins+ it seemed to kinda fix itself and would update every 5 seconds or so. But later I picked a new object to view and it went back to just constantly stacking all the frames on top of each other. Causing major star trails and horrible images.

What is going on? Even with an alt az mount I don’t think I should ever be seeing star trails at x256 or x512 should i?

Well done if you managed to follow what im trying to say lol!  

Sorry if its been asked before! I just don’t understand whats happening

[johnno]

shutter set at 1/50 and X256 = about 5 sec        X512= 10sec    X1024 = 20 sec.. that's how i think it is,  the more technical guys will explain it better.

[shirva]

Reading from the bible..no matter what camera is set to, it is constantly outputting full frane pictures at a clocked rate of 25fps or 30fps regardless of exposure setting or frame accumilation setting..you can set the frame rate down in capture device to say 1fps capture setting ,only one will be digitized every 25/30 pictures streaming from the camera output as a single image each second.

So if using x128 accumulation mode refreshes an image every 2.56 seconds and if capture rate is 25fps your movie file will contain 25 copies of the same accumulated image for each second of footage

Example

1 min recording containing unique images outout every 2.5s is(60s/2.5s capture rate =24 unique images

1 min recording captured at 1fps=(60s/1s capture rate=60 images)

A capture rate matching or closely matching the cameras refresh time is best......post more the morra..lol

[hoops86]

shutter set at 1/50 and X256 = about 5 sec        X512= 10sec    X1024 = 20 sec.. that's how i think it is,  the more technical guys will explain it better.

Thanks John. i knew my numbers were wrong in the example  

Cheers Davy i think i understand most of that lol. I was messing about with settings and tried to go more than 30fps but it said it couldn't go higher. Then i put it down to 1 and the image sucked so put it back to 25. What i dont understand is why my images are not refreshing every x amount of seconds. I must be doing something stupid. Just checked and its clear outside now! will see how i get on

[Dragon Man]

You aren't alone Chris.

Many people don't quite understand how an integrating video camera works.

Sens-up? Integration?

They are both the same thing.

I will try explain this in simple terms and avoid technical hoo-hah 

If a camera was to take say a 5 second image repeatedly, then it would be an Imaging camera, taking 5 second long images, and downloading an image every 5 seconds.

But how CCD Video works is that the frame rate is still 1/50th of a second, but the sensitivity of the chip is turned up and accumulates how ever many frames you tell it to (x256, x512 etc).

So, what the camera is actually doing is taking 256 1/50th second images at a much higher sensitivity.

SENS-UP = Sensitivity Up.

The camera will now store and stack 256 higher sensitivity frames each 1/50th of a second, totalling 5 seconds of high sensitivity stacked images.

So, the sensor doesn't stay 'open' longer. It's sensitivity rises.

Deep Sky Imagers will tell you that 1 frame at 20 minutes will contain only so much data. Not enough for a good image. So the Imager will take 10, 15, 20 images and stack them together for much more detail, better signal to noise ratio, accumulation of light, etc.

Notice he hasn't lengthened the exposure times, just taken more of them at the same exposure.

Same with Video.

Except he sensors we use are extremely sensitive. Turning up the sensitivity allows it to collect a lot of light (data) in only 1/50th of a second! Pretty amazing.

Add lots of those 1/50th of a second frames together into the Amp and Bingo, lots and lots of data.

You may have heard of the term 'Amp Glow'. The 'AMP, is like a storage tank of the images. All 256, 512 or whatever amount you select, accumulate in the Amp (imagine it as a memory chip). The more you shove in it the hotter the amp gets, which in turn heats the sensor near it and creates Amp Glow which appears like light seeping into the sensor.

That is what Cooling controls.

The Amp stores all the frames, and on-board circuitry internally 'stacks' them, then sends it on its merry way out of the camera, continuously.

While the first batch is being collected, stacked and sent, the 2nd batch is already accumulating, and as it is being stacked and sent the 3rd set is being accumulated, etc etc etc and on it goes, giving it the appearance of Video. Much like the old 'Flip Book' animations we made when we were kids.

I hope that makes sense. 

[johnno]

Looks like I've been doing it wrong?

 Ken, [ our Guru]  I use Gstar to capture and I just set it to 1000 frames and it whizzes through them in no time.  I've just set it up again and this time had a good look at the settings.

capture settings only go up to  256 and in the box it says 5.12 sec , so if i'm doing 512 do I input 10.24 sec in the box and leave it at 256 in the box, and what you said on another thread I can in theory go up to 1024 with the Sammy so do I input 20.48 sec in the box .... can I go higher still by changing the integration output rate?

So how many frames would you recommend to capture for processing later on?  I've just looked at the status countdown [another thing I've never bothered with before ] and at 256 @ 5.12sec @ 1000 frames takes 1hr 25mins, so this is where I must been going wrong as before 1000 frames would be done and dusted in about 20 sec.

Also what do you check in the video renderer box?..

cheers John

[hoops86]

Thanks a lot Ken! That helps me understand much better!

Could you try and explain what factors impact the amount of time it takes to refresh the image on my screen? I have noticed a big difference in the time its taking to refresh ranging from 10 secs to several minutes and i have no idea why. I assumed it would always refresh the image at x seconds depending on shutter speed settings.

Do things like focal length or visability or things like that affect it? Or if i was to remove my diagonal and put the camera right into the back of the scope would that impact the refresh rate? Or is there settings in the capture software i can change that would affect it?

Im just trying to understand why sometimes it will give me nice images that are getting updated every 10 secs or so.... compared to other times im seeing it update sooooooo slowly and produces horrible images on-screen with bloating and star trails everywhere.

[Dragon Man]

Looks like I've been doing it wrong?

 Ken, [ our Guru]  I use Gstar to capture and I just set it to 1000 frames and it whizzes through them in no time.  I've just set it up again and this time had a good look at the settings.

capture settings only go up to  256 and in the box it says 5.12 sec , so if i'm doing 512 do I input 10.24 sec in the box and leave it at 256 in the box, and what you said on another thread I can in theory go up to 1024 with the Sammy so do I input 20.48 sec in the box .... can I go higher still by changing the integration output rate?

So how many frames would you recommend to capture for processing later on?  I've just looked at the status countdown [another thing I've never bothered with before ] and at 256 @ 5.12sec @ 1000 frames takes 1hr 25mins, so this is where I must been going wrong as before 1000 frames would be done and dusted in about 20 sec.

Also what do you check in the video renderer box?..

cheers John

John,

I'm afraid you have me at a disadvantage.

I have never heard of any settings where you tell the camera to take 1000 frames.

I only know Sens-Up times as in Samsung and some other brands, and seconds as in Mallincam.

I also don't understand how or why you are setting the Sens-Up in the Gstar-ex software.

You should be setting it in the cameras menu.

I'm no Guru, just sharing any info I know 

[hoops86]

John,

I'm afraid you have me at a disadvantage.

I have never heard of any settings where you tell the camera to take 1000 frames.

I seen this in Gstar too btw! When you click the session button (i think). You could set it to either use a max time or max frames and just enter the amount you want. And there is also a integration drop down box that lets you select different speeds. I had no clue about any of it though so i just left them as they were

[Dragon Man]

Thanks a lot Ken! That helps me understand much better!

Could you try and explain what factors impact the amount of time it takes to refresh the image on my screen? I have noticed a big difference in the time its taking to refresh ranging from 10 secs to several minutes and i have no idea why. I assumed it would always refresh the image at x seconds depending on shutter speed settings.

Do things like focal length or visability or things like that affect it? Or if i was to remove my diagonal and put the camera right into the back of the scope would that impact the refresh rate? Or is there settings in the capture software i can change that would affect it?

Im just trying to understand why sometimes it will give me nice images that are getting updated every 10 secs or so.... compared to other times im seeing it update sooooooo slowly and produces horrible images on-screen with bloating and star trails everywhere.

Chris, the type of scope, focal length, aperture, weather etc have nothing to do with refresh rates.

However, your capture software and computer do.

If you have the integration set for 10 seconds (Sens-Up x512) the image should refresh every 10 seconds. If it doesn't, it sounds like the software or your Computer are lagging.

The most lag I have ever had is about 5 seconds, and that is Flash being slow sometimes when I am doing a live broadcast. Rare but it happens.

Single core computers can struggle doing too many things at once. Sometimes even Duo Cores have a struggle to keep up. But they usually mange.

Quad Cores make it a lot easier.

How many programs do you have running at the same time while your camera is capturing.

As an example, I have a Quad core Win7 and I run the Camera data in via Video Grabber, camera control out via Mallincam Control Software, WebcamMax to control image manipulation and broadcast window, run Night Skies network on Firefox using 4 lines of data ( NSN in, me send ing image back out, image coming back to me a s a broadcast, then out again with added settings as the final broadcast that viewers see, I also have Cartes du Ciel open and running, and I have PhotoShop open so I can save Screen Grabs.

No wonder I occasionally get lag 

[johnno]

ken I've set the sens-up in the camera @ 512 and in the session tab there is Capture Type. what I've been doing is to check manual capture .. maximum number of frames, I set this to 1000 like I said it would do it quickly. but now I've had a go at changing the other settings in Capture Settings  and 1000 frames is way to long, I was just wondering how many frames do you normally capture for processing 

[Dragon Man]

I seen this in Gstar too btw! When you click the session button (i think). You could set it to either use a max time or max frames and just enter the amount you want. And there is also a integration drop down box that lets you select different speeds. I had no clue about any of it though so i just left them as they were

Ahhh, I know where you are seeing this.

I have never used the 'Session' area. But the chap who wrote the software is a friend of mine and sometimes comes to our Astronomy camps (unfortunately he won't be at our next camp which is this weekend or I could have asked him more).

He explained it to me, as best as I could understand, about 2 years ago, and I posted some of the info into Cloudy nights Forum, then he came in and added more info. But I myself never used it.

When I use the Gstar software I only use the 'Camera' setting under 'Tools'. then 'Properties', then 'Video Proc Amp'.

I set the camera settings in the camera, but set Brightness, Contrast, saturation etc in the 'Video Proc Amp' window of Gstar software. 

[Dragon Man]

ken I've set the sens-up in the camera @ 512 and in the session tab there is Capture Type. what I've been doing is to check manual capture .. maximum number of frames, I set this to 1000 like I said it would do it quickly. but now I've had a go at changing the other settings in Capture Settings  and 1000 frames is way to long, I was just wondering how many frames do you normally capture for processing 

John, I don't capture for processing.

As I just posted above, I don't use 'Session' in Gstar.

I only do Live viewing or Broadcasting. The Session area is more for Imaging.

[johnno]

I seen this in Gstar too btw! When you click the session button (i think). You could set it to either use a max time or max frames and just enter the amount you want. And there is also a integration drop down box that lets you select different speeds. I had no clue about any of it though so i just left them as they were

that is exactly what I've been doing, but now I've had a closer look at it , just need to experiment more with it, I think what as been happening is it was taking 25 to 30 fps so that is why it was done with quickly, So now if I set it to capture 20 frames @ X256 it will do it 3min .25sec taking a single frame @ 5.12sec so if I go to 512 sens-up it will double the time, but is 20 frames enough for processing?

[hoops86]

Chris, the type of scope, focal length, aperture, weather etc have nothing to do with refresh rates.

However, your capture software and computer do.

Ok thanks thats exactly what i needed to know! Time to bin the netbook and start using the mac book pro i think

[johnno]

Thanks Ken , you cleared another thing up, session mainly for processing

[shirva]

Read somewhere about 700,,cant remember if it included flats/darks/lights...think it was in the book by massey and quirk on astro video...brilliant book...davy

[DoctorD]

Hi Ken,

I'm not sure it works how you say - I think that the SENSUP does hold the shutter open for an extended time, the post processor or DSP sends the same picture out every 1/25 or 1/30 of a second as required by the video standard whilst the next frame is building up.

Mind you this is only my opinion based on personal experience.

Paul

[hoops86]

Ive tested it on the mac and its exactly the same when i go up to x512. Sometimes im getting very clear crisp images refreshing every 10 secs or so. But most of the time its lagging and taking minutes to refresh giving me horrible images i also noticed that it starts to lag even at x128 and x256 but i can live with it as it does update...just slower than normal.

One thing to note is there is no lag at all when using the menu of the camera! I can see my movements through the menu instantly. So im thinking it must be the camera thats the cause of it. Not the software, cables, capture card or spec of laptop.

No clue what to do now.

[shirva]

I had the lag in four channel usb grabber but wasnt camera..so grabber or software culprit,,,davy

[shirva]

Hi paul,,not heard from you in a wee while,,,all ok i hope..or are you just keeping head down,,,davy

[Macavity]

I see it as an "experimental science"? lol. In practice e.g., I "time" the (integration) screen updates in (real) seconds and set the (GSTAR) event writing rate to correspond approximately with that. I still wonder vaguely about (de-)interlacing and stuff. I feel I have a reasonable handle on things. But one can be forgiven for occasional doubts... Even "authorities" can be wrong by a factor of TWO or so...

[Dragon Man]

Hi Ken,

I'm not sure it works how you say - I think that the SENSUP does hold the shutter open for an extended time, the post processor or DSP sends the same picture out every 1/25 or 1/30 of a second as required by the video standard whilst the next frame is building up.

Mind you this is only my opinion based on personal experience.

Paul

Ahhh, OK Paul, thanks for that.

I'll have to do some more reading up. I know that there are 2 different types of Video methods from different Brands. Maybe I have my wires crossed (wouldn't be the first time) 

The way I explained it is the way It was explained to me by one of the Manufacturers.

The way you explain it though makes more sense about how it is possible.

My brain hurts 

[Carl Reade]

Hi all here is a link which maybe useful,

 http://www.cctvdynamics.com/what-is-sense-up-an-introduction-to-and-explanation-of-the-most-misunderstood-security-camera-feature.html

Cheers Carl

[hoops86]

Cheers for the link Carl. Interesting read! i think im slowly starting to get my head around this stuff

i think this statement from the link above is important in relation to the poor images at high integration:

"When a sense-up camera adjusts the shutter speed it is essentially allowing so much time for light to enter the camera between frames that it becomes impossible to send 30 frames every second. This is fine for static images where things don’t move very much. Each frame is the same as the next, so there is no movement in which we can see the evidence of the screwed up syncing. This changes altogether when there is a moving object in the image, and the faster the object moves, the worse the image appears. What you will see in this situation is best described as extremely blurry images. The DVR will capture up to 15 frames for every one frame the camera sends, and they will not be synced up properly."

Sounds like my camera and software are not synchronising correctly......