New Imaging Source cameras

[michaelmorris]

I'm started saving for the new DMK21 yesterday.

[michaelmorris]

Couldn't wait any longer, one DMK21AU618 ordered from FLO.

[FLO]

It appears some USB controllers built into PCs might not be able to handle the data rate required for 60 FPS. If you have USB3 (USB3 is better even in USB2-mode) you should not experience a problem. The same might be true of some Firewire connections but we don't have enough feedback yet to comment.

Thanks to Neil Philips for bringing this to our attention

Steve

[Ant]

Just a quick note that might help someone.

I had (and still have) a DMK21 - with my old laptop it would work quite happily with one of the USB ports, but refused to work for more than a few frames (then it just froze the whole laptop) on the other USB port.

As far as I could tell the laptop thought that these two USB ports were exactly the same.

So if you have issue with data transfer speed with one USB port it is well worth trying it on another before you start looking for other issues / fixes.

Cheers

Ant

[riklaunim]

Some laptops may have underpowered USB ports (or USB 1.1 in older laptops) and that may cause problems. Solution could be a powered USB hub (but it also gives mixed results).

[blackparticle]

I'm a big fan of hi-res / hi-mag lunar pics and those shots with the DMK72 look superb.

Looking through the SGL lunar gallery it seems it's not something that is commonly attempted.. probably because it's so difficult. I've tried it with an SPC900 but at 640x480 it's hard to get decent images as they usually end up quite tiny as I have to scale the image downwards to account for any blurring.

I'd love to see a shoot-out between that and the Starlight Xpress Co-Star.

[neil phillips]

Thought i would join in this conversation, thanks for starting a conversation about this Steve. as far as i was aware this pc had usb2 ports i think it was released around 2008, heres my hardware specs i could find

HP Pavilion a6544.uk Desktop PC Product Specifications HP Pavilion a6544.uk Desktop PC - HP technical support (United Kingdom - English)

The odd thing is the live feed is not collapsing after i press record, but when i tell IC capture to change to 60 fps. Does this sound like a computer or camera problem i wonder. clearly it would be good if i could try on another pc. Freddy williams is running one of these no problem, i wonder what computer specs hes got, i will message him

Also space cowboy is running at 60 fps, with a machine ( laptop ) that seems less well speced than mine ?

[riklaunim]

DMK72, QHY5 and Costar use CMOS sensors with rolling shutter. They do have a lot of pixels but are harder to use for imaging as they give "fixed pattern noise" common for mid-level CMOS sensors. That's why DMK72 isn't officially in the Astro category. Those cameras do work if you try hard (you can find QHY5 Moon shots on the net), but not as good as CCDs - now leading ICX445 and also ICX204/205 in DMK31 and 41 - for Lunar and Solar. The difference is price. You will pay much less for the CMOS than for a ICX445 ExView HAD camera (or Costar/QHY5 cheaper than Lodestar). If all goes well exmor CMOS devices will keep the low price while reducing noise issues noticeably.

And also note that when you have a big frame - collection a lot of frames takes some time and then you end up with a very big AVI file that may be very time consuming to stack :] Thats why those DMK72 Moon shots from Germany used slight ROI.

[blackparticle]

Cheers for that info Rik.

I can see how for multi-purpose solar system imaging they might not be as good as a CCD but for dedicated lunar work, I think the trade-off between sensitivity and image frame size could be worth it.

The great thing about imaging the moon is that it is (almost) permanently fixed in its rotation so you don't really have to worry about keeping the AVI's down to 3 minutes or so like you do with Jupiter or Saturn. I haven't calculated how fast the shadows from the sun move across the surface but I imagine at the magnification one could achieve with an amateur scope then a 10 - 15 min AVI isn't going to reveal significant differences.

I agree that this does lead to some massive files and increased processing time but as always, it's the results that count.

The fixed-pattern noise aspect of the CMOS can be minimised by slightly moving the target image around in the frame as you image.. as well as applying darks obviously. You could even rotate the camera 180deg and shoot a second video then stack them together.

As old Selene kind of screws up DSO imaging for 1 week out of every 4 then it certainly makes sense, to me, to have something specifically for the task.

[riklaunim]

You can even hunt exoplanet transits with a DMK (there is a PDF on the net) it's all about trying hard and processing. That DMK72 has the advantage of ROI (and binning) vs all other DMKs. The pattern noise & bad stuff can be limited by dark and flat frame + natural slight object drift in the fov (as it's unguided ).

[Whippy]

FWIW, I've had the same issues with my camera (same sensor) in testing and I've found that it's more down to the machine just not having enough performance rather than the USB controller itself. Even something as simple as adding more memory helps a little but to me, it's just that slower machines just can't cope with 60fps in 8 bit and it's a darn sight worse if you want 12 bit mode at 30fps (can the imaging source cams do that?). XP is a huge bonus over Vista or W7 too .

Tony..

[neil phillips]

FWIW, I've had the same issues with my camera (same sensor) in testing and I've found that it's more down to the machine just not having enough performance rather than the USB controller itself. Even something as simple as adding more memory helps a little but to me, it's just that slower machines just can't cope with 60fps in 8 bit and it's a darn sight worse if you want 12 bit mode at 30fps (can the imaging source cams do that?). XP is a huge bonus over Vista or W7 too .

Tony..

Thanks for the feedback Tony, but lets be clear here, are you saying that you had problems recording at 60 fps. or that as soon as you changed the setting from 30 fps to 60. the camera shuts down and is no longer sending a live feed at all. Because thats whats happening here. i understand there can be problems recording at 60 fps. But im not saying that. im saying the camera completely shuts down a live feed, just because i highlight 60 fps in the tab ?

[Whippy]

From my experience Neil, I've had problems just having the camera running at 60 fps on machines, preview and recording. I've tested it on 4 different machines (2 netbooks running W7, 1 laptop running XP and 1 desktop running Vista) and the lowest spec machine has the most problems. I'm trying to sort out a dual boot setup for one or more of these so I can make definate comparisons but shovelling that amount of data to my mind makes it a performance issue more than anything else. What you've mentioned sounds more like a hardware/software bug than anything. Have you tried it on more than one machine/OS?

As I mentioned I don't have the IS camera or use the software so I can't compare, I can only go on from what I've found which to my mind is that these cameras using a USB bus are running somewhere near to it's maximum performance and in a lot of cases, the nearer you are to the maximum, the more chance things will go wrong. On the software for my cam, I can adjust the framerate or let it run as fast as it can which theoretically is 60fps in 8 bit mode or 30fps in 12 bit mode which for the two netbooks is nowhere near that.

Tony..

[neil phillips]

Thank you again Tony From what your saying it does sound like im experiancing some kind of hardware software bug.

This doesnt appear to be a performance issue i dont think. The specs of my computer may cause capture problems at different frame rates. lots of people are experiancing that. but this issue appears different, in that just highlight a setting is causing the camera to crash.

If it was a performance issue i would expect this to be happening after recording has started, i could be wrong. But any performance issues ive ever had, have always been after recording started. not highlighting settings.

Steve ( FLO ) and hes team are being extremly helpful on this. And hes told me we will work the problem, one way or another. which is one of the reasons, i will continue to purchase equipment from him, the after care service is second to none. Above and beyond what you expect from any dealer.

For now im going to try to get some images at 30 fps, and work with steve. to get the camera operating at full 60 fps. I shall be posting images up at some point showing what this sensor, and this particular camera can do.

I know images are everywhere on the net with this ccd , and even some are appearing now, with this exact camera.

But it will be interesting seeing what we can get in the UK with this camera, and with fairly budget equipment, compared to say C14 s

So say tuned for that. and possibly a few images and review on FLOs website. All new equipment can suffer bugs. and we just need to figure out what the cause is, if others get the same problem we will then know the cure.

If anyone has any guesses on what is likely the problem. please continue to post. it could be just my particular computer. But im leaning away from that, and more a camera bug of somekind. comparisons with another camera and computer. will be forth coming. stay tuned for findings in the next couple of weeks

[blackparticle]

Have you considered the quality of the USB cable? Even just changing the settings to 60fps will start to stream data in preview mode. It maybe that the data stream is getting totally corrupted.

I've never come across anything where the USB cable affects performance like you are describing but with the exception of data transfer from an external HDD, don't have anything that would come close to maxing-out the USB2.0 interface.

I imagine data-transfer from a HDD also employs error correction whereas live streaming video may not? (Hence why certain frames are dropped)

Just a thought.

[Whippy]

Fingers crossed Neil . Therein lies the problem though, manufacturers can't possibly take into account every variation of machine so early adopters iron out the kinks to help stablilise the setup. I hope you'll be up and running at 60fps soon, I'm sure it'll make the world of difference on Jupiter.

Tony..

[Space Cowboy]

Like Neil I am having problems with my DFK 21AU618.AS. I can't achieve more than 30fps in ICC Capture using any codec. Firecapture does produce 45fps using y800 and 55fps with ROI. I can't use the BY8 codec under any frame rates...as soon as it is selected the capture software crashes saying there is no live feed. This happens with ICC capture, firecapture and sharpcap.

My other concern and I wonder if anyone can clear this up is something I've read since receiving the my DFK 618.

I thought the only difference between a DFK and DBK was the dfk having an inbuilt IR blocking filter.

From what I've now read it seems there are more differences. For one its been suggested the in-built IR filter is optimized for daytime use and not suitable for astrophography. Secondly this passage below says the dfk does not produce RAW data like the DBK meaning slower frame rates:

"Color CCDs contain a Bayer color microlens matrix - each pixel features a green, red or blue filter only, colors are arranged in the Bayer pattern. The missing two colors for each pixel are calculated from adjacent pixels having the required color. The DFK models do this debayer calculations in the camera, the camera output is complete RGB for each pixel. The DBK models output the raw bayer data to the host machine where the debayer calculation will be done instead of in the camera, the main advantage is speed: Only one third of the full RGB data has to be transfered and higher frame rates are possible."

taken from this website :

http://www.astro-imaging.de/astro/Im...04.AS_p01.html

I've all ready shown this to a leading imager who thinks its probably nonsense.

[neil phillips]

Dont take my word on this Stuart, i said i could be wrong. but if this info is correct, then you will not be able to use debayer on reg6, something i experimented with, and had good results with using a DBK, I hope someone can tell you for sure if this is indeed the case.

I have just asked permission to show specs of a computer of a astronomer whos using the DMK 618. im sure he wont mind me saying the specs of the computer are far inferiour to mine. Hes of the opinion its the camera not my setup. after i saw hes computer specs, at the moment im inclined to agree. though i cant rule out some kind of bug or clitch on my particular system. though it seems unlikely. If it is the camera, it must be bad luck, as many are being sold without problems

[riklaunim]

If old DBK and DFK worked at 60 FPS then I doubt that they differentiated DFK and DBK now. Maybe try also with huffyuv?

There is also some software update: http://www.cloudynights.com/ubbthreads/showflat.php/Cat/0/Number/4730729/page/0/view/collapsed/sb/5/o/all/fpart/1

[neil phillips]

Here the specs a astronomer whos getting great results with this usb camera has.

Dell= Geniune Intel CPU T2300 1.66GHz. HD= 80 GB. Ram Memory = 1 GB.

Assus= Geniune Intel CPU 2400 1.83GHz HD= 230 GB. Ram Memory = 1 GB.

They are not even half as fast as yours....

Hes quote at the end

There has been mention to me The AMD chipsets may be problematic. but on firewire only dropped frames was a issue, needing a clean hardrive, properly defraged. other than that no problem. chipset seemed fine on firewire.

[Space Cowboy]

Thanks Neil, yeah I was careful not to quote the imager by name

Thanks for the codec tip Rik. The info I quoted from that website was referring to the old DFK I'm not suggesting the 618 version is differently setup just assuming if the old dfk differed to the old dbk then the new one would too.

[Space Cowboy]

This is the thread on CN where I found the info about the DFK being different to DBK :

Posts 4 & 6

CN FORUMS ARCHIVE Telescope Reviews: Firewire DFK 21AF04.AS vs. Lumenera Skynex 2-0

[hayju01]

Just got email to say my DFK 21AU618.AS Colour has left FLO

Now I just need to tell the misses

[Space Cowboy]

Read the CN thread Rik mentioned and changed exposure to 1/61 (did not realise it could be changed by one digit) BINGO! 60 fps with y800 every time!

Ran the mono y800 avi through Registax GB Debayer and BINGO! Colour image!

So I guess this means the stuff about DFK not producing RAW data is not true!

[neil phillips]

That sounds like good news Stuart how did you change the exposure by just 1 increment firecapture ?

good news its debayering properly then. you are certain its producing a colour debayered image ?

It seems i was wrong in my belief ( and other planetary imagers i spoke to) that this camera is faulty, it appears certain AMD chips my have a bug with it. But not all. I travelled to my eldest daughters and tried her laptop. and surprise surprise it worked straight off. still had the problem of setting 1/76th exposure to get it to run at 60 fps. it dropped 300 frames. but the frame count was pretty respectable considering the machine isnt a top spec model. So it seems if this problem occurs a work around may be to replace some usb slots with a pci card. yet to be determined.

But i can say i was surprised by this finding, especially as it was a AMD that worked. ok it wasnt the same AMD chip as mine. but AMD its a odd digital world out there