From: "steve_chambers_1"Date: Thu Feb 7, 2002 1:01 am Subject: Re: ExView CCTV news... Hello, I would like to write a few words about this camera. As most of you will know Jon has been working with the 1004x for some months. Jon developed a manual control of the shutter and has posted some images including a striking m42 and developed software. My own input to the project has been relatively minor. As Jon has said the modified video camera should be able to offer, Computer controlled exposures, On-chip amp control Full vertical resolution this has not been easy!! Possible self-guide Probably the biggest factor though is the fact that the camera uses sony's fabled ex-had chip. To put this into context. The vesta uses a had chip. The ToUcam uses a `new structure had' (I think), which is about twice as sensitive. The ex-view chip is at least twice as sensitive again. Moreover the webcams use primary filters on the chip which reduce the sensitively by 2/3 thirds. Also the chip in the 1004x is larger than the webcams. Result. Sensitivity wise this camera, has the potential to be, on a par with `ANY' of the commercial ccd imagers from SBIG, Apogee, Starlight etc. If you have a interest in objects from the Abell catalogue this is what you have been waiting for :-) Steve C --- In QCUIAG@y..., Jon Grove wrote: > Hi Group, > > Just to whet your appetites, I thought you might be interested to hear that > Steve Chambers and I have been working on some modifications to the 1004X > ExView monochrome CCTV camera, and we have made several breakthroughs. > > I have a prototype setup which allows: > > * Full-frame 640x480 capture. > * 'Autoguide mode' (640x240 live video + simultaneous 640x240 integration). > * 100% capture success (no missed exposures). > * Amplifier disable. > * Peltier cooling. > > Last night I took a (somewhat out of focus) cooled, 120second, amp-disabled > 640x480 image of a corner of my room. This was the first time I'd tried out > all the new features simultaneously, and I was pleased at the success. There > are a few hot pixels, in part because the cooling was not 'full on' and I > haven't yet applied thermal compound to all the joints in the cold finger > assembly. But as a proof of concept - well, judge for yourselves. The room > was lit by 3 or 4 LEDs, and a digital clock on the other side of the room. > The shadows on the wall and the highlight to the right were cast by a small > red LED on the PSU for the Peltier. The dark left-hand corners are > vignetting because the CCD was not centralised properly beneath the lens. > The image is completely unprocessed other than to rotate it 180 degrees (the > camera was upside-down at the time!) > > At present my setup is manually controlled, but I'm confident that it will > be PC controlled (via the printer port) in the near future. > > When we've finalised a circuit diagram and got some software sorted out, > I'll let you all know. Watch this space... > > Clear skies, > Jon Grove. From: "elburro51" Date: Mon Feb 11, 2002 10:18 am Subject: ExView + Firewire,YUV 444 =request for firewire webcam chipsets Hi All, First thanks to the camera gurus (Steve, Jon, Sylvain, Etienne... etal) for all your work in giving us access to cheap, long exposure (+guiding) cameras. (+ not forgetting the software guys as well) This is just a wild idea put up into the ether with little (almost none) hardware checking out. Considering...... (A) Fire wire = much faster bridge than USB, therefore perhaps less initial compression of the raw CCD image before transmission to computer. Also the firewire webcams are able to do YUV 444 format And reasonably affordable (All 5 priced around $US90-100 street) (B) Sony ExView Chip = great sensitivity compared to webcams (HAD CCD chips in good ones) i.e.(i.e. Jon's RF Concepts B&W board camera) But requires (limited by?) conversion to computer format i.e. D/A then A to D via external grabber card. (C) Steve's and Jons mods to CCD support chips for long exposure. (thanks again guys) (D) From Sony Ex View info on RF Concepts PDF doc "since it has same pixel count, same pin config, same drive specs as conventional products it can be used as a direct replacement" Sony 1/4in 3.8k pixel CCDs ICX068AL/069AL, ICX208AL/209AL Questions Arising:- 1. Is there any of these firewire cameras that could have an Exview chip graft to give us the best of both worlds.(or even graft to USB mod exposure sorted webcams ie Vesta/Toucam Pro) 2. Sources of "cheap Exview 1/4in chips (B&W and color) 3. What CCD/driver chips do these firewire use? So any one out there with any of these firewire cameras that they can post the CCD/Driver chipset numbers? /internal photos Is this much more complicated that I think? Cheers Greg Below is links to the five webcams I can find on the market:- IREZ (2 cameras) http://www.irez.com/products/kritterdigital.html http://www.irez.com/products/stealthfire.html Both ¼ HAD 350Kpixel (sony?) CCDs advertised but some QCULAG site questioning of this claim (CMOS? Image chip) FIREWIRE DIRECT WEBCAM http://www.firewiredirect.com/firewire/products/cam/Cam_comparison.sht ml http://www.firewiredirect.com/firewire/products/cam/specs.shtml ¼ HAD 350Kpixel (sony?) CCDs TI driver chipset? IBOT http://www.orangemicro.com/ibottechspecs.html ccd chip? TI chipset? ADS PYRO http://www.adstech.com/products/PYRO1394WebCam/specifications/PYRO1394 WebCamProdSpec.asp?pid=API-200 They advertise "PYRO 1394 Web Cam has ZERO compression & utilizes a high-quality Sony(r) CCD" (3.3k ¼ in) But does seem to get mixed reviews for picture quality and needs extra power if connected to some laptops From: "gcne67" Date: Mon Feb 11, 2002 3:59 pm Subject: Re: ExView + Firewire,YUV 444 =request for firewire webcam chipsets [snip] > Below is links to the five webcams I can find on the market:- [snip] There is also the Unibrain webcam: http://www.1394store.com/eshop/product.asp?dept%5Fid=51&pf%5Fid=5310 From the spec sheets it appears to be the same as their "machine vision" camera. http://www.1394store.com/eshop/product.asp?dept%5Fid=51&pf%5Fid=5320 They use the ICX098AK CCD (same as Vesta), the fellow doing the Linux drivers for IIDC compliant cameras just tested the "machine vision" version and told me it was useless. Shutter exposure seems to do nothing. Low light performance was hopeless ("sucks" to be exact) and needed maxium gain just for an indoor shot. No 640x480 RGB mode (i.e. to avoid compression). Lower resolution is based on skipping pixels, rather than binning, amongst other problems. Replies Auth From: "steve_chambers_1" Date: Wed Feb 13, 2002 1:06 pm Subject: Re: 1004X board Hi Adam I have both a board camera without audio and a camera in a little square case with audio both from rf concepts. There is little diffrence between then. The components are slightly more tightly packed on the latter. Steve C From: "vilarestil" Date: Wed Feb 13, 2002 9:55 am Subject: 1004X board Looking forward to the mod details for the 1004X board by Jon and Steve. Any idea of timescale as this one sounds exciting?! RF Concepts are now offering the 1004XA camera board which is supposed to be the same thing but with audio. Does anybody know if this is the case? Question for Jon or Steve, did you get your original prototype board from RF Concepts or is there another UK supplier of these excellent cameras? All the best for now, Adam From: Jon Grove Date: Thu Feb 14, 2002 8:11 am Subject: RE: [QCUIAG] Re: 1004X board Hi folks, Sorry if it seems I'm keeping you in the dark (although that's not such a bad place for an astronomer). I'm just tidying up the design of the mod at the moment, and sorting out some software to drive it. I want to be sure that when I do publish it I don't have to tell people to redo all their soldering because of a design flaw! Shouldn't be long now... Jon. From: "rjsdotorg" Date: Thu Feb 14, 2002 7:57 am Subject: Re: 1004X board Looking forward to the mod details for the 1004X board by Jon and Steve. Any idea of timescale as this one sounds exciting?! I agree, I bought one and use it at 60fps, for now. RF Concepts are now offering the 1004XA camera board which is supposed to be the same thing but with audio. Does anybody know if this is the case? That is the new model which I have. I'll post pics at: Pictures and More in a few minutes. Question for Jon or Steve, did you get your original prototype board from RF Concepts or is there another UK supplier of these excellent cameras? I think that he said his was from RFC. The only other links I have ever found are: 1004X 1004X The chips are well documented, but my electronics skills are rudimentary. Ray From: Jon Grove Date: Fri Feb 15, 2002 1:41 am Subject: ExView fullframe first celestial light Last night there were only thin whispy clouds in my area, and so I decided to try the newly 'fullframe' modified 1004X camera on some real astronomical objects. It was rather windy and everything was wobbling around a bit, but I thought I'd have a go anyway. I took a few exposures of the M42/M43 area, ranging from 6 seconds down to 0.5 seconds, and I also took some of the Flame nebula, at about 6 seconds. The sky was a bit bright (streetlights reflected off the thin clouds) so longer exposures didn't really make much sense. I tried M81 and M82 but the light pollution was too bad in that direction. M51 was not visible at all. Anyway, here is a mosaic of the M42/M43 region, and my Flame nebula (overprocessed to combat the light pollution). Clear (and still) skies, Jon. From: "vilarestil" Date: Fri Feb 15, 2002 4:51 am Subject: Re: ExView fullframe first celestial light Excellent images for your prototype, Jon! Going to order a 1004XA board from RF Concepts today!! Got my soldering iron heated up and ready to carry out the mods as soon as you publish them - hi!! All the best, Adam From: "dehahne" Date: Fri Feb 15, 2002 7:49 am Subject: Re: ExView fullframe first celestial light Jon, What set up did you use to capture the M42/43 image (i.e. scope specs, focal reducer, etc)? I'm currious to see what it takes to get the kind of field of view you captured. I guess I would also need to know the CCD dimensions on the 1004X as well. Thanks. Dave Hahne Poquoson, VA USA From: Jon Grove Date: Fri Feb 15, 2002 9:12 am Subject: RE: [QCUIAG] Re: ExView fullframe first celestial light Thanks, Richard. Yes, the camera uses the standard Sony chipset - I suspect the mod will apply to the PC164C with only minor changes, if it uses the same or similar chipset. It looks as though the PC164C has a few controls on it (the 1004X has none!) so the PCB is no doubt laid out differently, but I'd guess that the circuitry probably resembles the Sony suggested circuits which they publish in their spec sheets for the ccd and timing generator chips. Basically, if the camera uses the ICX2463R timing generator then the mod should work fairly easily. Jon. From: Jon Grove Date: Mon Feb 18, 2002 1:55 am Subject: ExView, more results Hi all, Friday night was clear, and despite being rather tired I decided to make the most of it. I was glad I did, as I managed to catch some good images using the newly-completed full-frame 1004X mod. The amp-off feature greatly improves the useability of the camera, allowing much deeper exposures without the image becoming swamped by electroluminescence. I took exposures of between 6 and 45 seconds or so for the attached images - unfortunately in my excitement I forgot to note exact timings! Most of the images are mosaics/stacks of a few individual frames, made using a piece of software I've developed myself. It allows you to manually align images and then automatically 'snaps' them to the best fit, adjusts brightness levels, removes hot pixels and 'feathers' edges of images to produce a fairly seamless composite. When it's a bit tidier I'll make it available to the group if there's any interest. Anyway, here are the images - M51 is my favourite aesthetically, because it's such an interesting shape. M81 was harder to capture so is my favourite from a technical viewpoint. Oh yes - the scope is a 6" f/5 Orion Europa. All the images were taken with an IR blocking filter, which seemed to help reduce skyglow a little. Jon. From: Jon Grove Date: Mon Feb 18, 2002 8:15 am Subject: RE: [QCUIAG] ExView, more results Thanks Steve! For those who are interested, the modified CCTV camera requires a video capture card in the PC (preferably not a USB device since that will use compression), and it needs to be able to capture fairly fast. I use a Hauppauge WinTV PCI card. If you can grab 25fps at 640x480 then it'll definitely work. If not, then I'm not yet sure whether the capture technique will guarantee 100% successful captures. Basically the software needs to be able to tell the camera to stop integrating and then guarantee to 'see' the next frame that arrives. It may be possible to synchronise it on a slower PC - I just haven't tried it because I've only got one computer! Jon. From: "rjsdotorg" Date: Mon Feb 18, 2002 11:31 am Subject: Re: ExView, more results It may be possible to synchronise it on a slower PC - I just haven't tried it because I've only got one computer! I have looked at most of the specs for the Sony chips (I have the NTSC model of the 1004x), and noticed that one of the features of the driver chip is the external sych options. With further review and maybe some help from others, a suitable sych could be provided from the PC, with the shift synch etc. provided by the chip-clock. I was most interested in the sych because it may enable us to read the CCD output into an extenal 16-bit A/D converter by intercepting the signals directly from the chip, and then sending over a port, a la the cookbook cameras. Some day. I've captured images with a 135 lens, but IR focus is a problem and the images are not worthy yet. :( The scope waits for better weather. From: "wlbehrens" Date: Mon Feb 18, 2002 12:07 pm Subject: Re: ExView, more results Now if I could just get the details on the mod :). I've got 2 of these cameras in my possision as well as a PYRO web cam (gona put the SC mod to it, should be interesting as it has a firewire interface). William --- In QCUIAG@y..., Jon Grove wrote: Thanks Steve! For those who are interested, the modified CCTV camera requires a video capture card in the PC (preferably not a USB device since that will use compression), and it needs to be able to capture fairly fast. I use a Hauppauge WinTV PCI card. If you can grab 25fps at 640x480 then it'll definitely work. If not, then I'm not yet sure whether the capture technique will guarantee 100% successful captures. Basically the software needs to be able to tell the camera to stop integrating and then guarantee to 'see' the next frame that arrives. It may be possible to synchronise it on a slower PC - I just haven't tried it because I've only got one computer! Jon. -----Original Message----- From: Dr. S.J. Wainwright [mailto:astrabio@n...] Sent: 18 February 2002 15:34 To: QCUIAG@y... Subject: Re: [QCUIAG] ExView, more results Jon, All of the images are good, but the M81 is the FIRST good image of this object ever produced from the group. It is very hard to capture the faint spiral arms. I can't wait to try it for myself. Best wishes Steve Dr. S.J. Wainwright Hi all, Friday night was clear, and despite being rather tired I decided to make the most of it. I was glad I did, as I managed to catch some good images using the newly-completed full-frame 1004X mod. The amp-off feature greatly improves the useability of the camera, allowing much deeper exposures without the image becoming swamped by electroluminescence. I took exposures of between 6 and 45 seconds or so for the attached images - unfortunately in my excitement I forgot to note exact timings! Most of the images are mosaics/stacks of a few individual frames, made using a piece of software I've developed myself. It allows you to manually align images and then automatically 'snaps' them to the best fit, adjusts brightness levels, removes hot pixels and 'feathers' edges of images to produce a fairly seamless composite. When it's a bit tidier I'll make it available to the group if there's any interest. Anyway, here are the images - M51 is my favourite aesthetically, because it's such an interesting shape. M81 was harder to capture so is my favourite from a technical viewpoint. Oh yes - the scope is a 6" f/5 Orion Europa. All the images were taken with an IR blocking filter, which seemed to help reduce skyglow a little. Jon. From: Jon Grove Date: Tue Feb 19, 2002 1:11 am Subject: RE: [QCUIAG] Re: ExView, more results The problem with capturing on a slower PC isn't so much synching the PC with the camera, since the capture card does all that. It's more a question of whether, in a system where not all the frames are being captured, we can stop the integration just before the card captures one of the frames that we know we will see. Reading the CCD output to a 16bit A/D might be an expensive option unless you want to slow it down a lot - the data is clocked out at a pretty high rate. 640*480X25fps = nearly 8M samples per second. If you slow it down, you lose the real-time advantage of video. You could have it switchable I guess. Perhaps a more practical option might be to use a cheaper 8bit A/D, which could convert at video rate and store in a RAM directly. The RAM could then be used to repeatedly send the captured frame as a video signal (a la Mintron), or be read out digitally at a reduced rate via a parallel port or something. 8bits would probably give better quality than is currently achievable using a capture card, since there would be much less opportunity for noise to creep in before digitisation. You could bypass all the AGC and gamma adjusting circuitry. Given the noise levels in the cheap CCDs used in these cameras, I would think going to 12 or 16 bits might be overkill anyway. If you want higher bit depth, take more images and sum them. Just my 2 penn'orth, Jon. From: Steven Jindra Date: Tue Feb 19, 2002 6:58 am Subject: RE: [QCUIAG] Re: ExView, more results Jon How much data would be lost if the live video is recorded to a Hi-8 camcorder or in my case a Digital-8 then later reviewed frame at a time to capture the high resolution, long exposure frame(s) to computer through a capture card? Of course the computer program would still be used to control the capture length. Steven Jindra --- Jon Grove wrote: The problem with capturing on a slower PC isn't so much synching the PC with the camera, since the capture card does all that. It's more a question of whether, in a system where not all the frames are being captured, we can stop the integration just before the card captures one of the frames that we know we will see. Reading the CCD output to a 16bit A/D might be an expensive option unless you want to slow it down a lot - the data is clocked out at a pretty high rate. 640*480X25fps = nearly 8M samples per second. If you slow it down, you lose the real-time advantage of video. You could have it switchable I guess. > Perhaps a more practical option might be to use a cheaper 8bit A/D, which could convert at video rate and store in a RAM directly. The RAM could then be used to repeatedly send the captured frame as a video signal (a la Mintron), or be read out digitally at a reduced rate via a parallel port or something. 8bits would probably give better quality than is currently achievable using a capture card, since there would be much less opportunity for noise to creep in before digitisation. You could bypass all the AGC and gamma adjusting circuitry. Given the noise levels in the cheap CCDs used in these cameras, I would think going to 12 or 16 bits might be overkill anyway. If you want higher bit depth, take more images and sum them. Just my 2 penn'orth, Jon. -----Original Message----- From: rjsdotorg [mailto:yahoo@r...] Sent: 18 February 2002 19:31 To: QCUIAG@yahoogroups.com Subject: [QCUIAG] Re: ExView, more results It may be possible to synchronise it on a slower PC - I just haven't tried it because I've only got one computer! I have looked at most of the specs for the Sony chips (I have the NTSC model of the 1004x), and noticed that one of the features of the driver chip is the external sych options. With further review and maybe some help from others, a suitable sych could be provided from the PC, with the shift synch etc. provided by the chip-clock. I was most interested in the sych because it may enable us to read the CCD output into an extenal 16-bit A/D converter by intercepting the signals directly from the chip, and then sending over a port, a la the cookbook cameras. Some day. I've captured images with a 135 lens, but IR focus is a problem and the images are not worthy yet. :( The scope waits for better weather. From: Jon Grove Date: Tue Feb 19, 2002 7:08 am Subject: RE: [QCUIAG] Re: ExView, more results Steven, That's definitely a viable technique - I used to record to a domestic VHS recorder because my PC is a bit bulky to keep taking outside. You get a bit of noise introduced, and some resolution is lost as well. Not sure how much better a Hi-8 or Digital-8 would be. Also there is sometimes some slight jitter between one frame and another, which is not a problem when aligning and stacking, but can make subtraction of hot pixels a bit more involved. And finally, on VHS at least, I used to find that bright stars had dark streaks to the right of them. But you do get the chance to record loads of stuff without running out of disk space, and you can resample it several times to try to reduce noise introduced after playback. And you have a permanent record of how long the exposures were, in the time between the brief flashes of 'image'. In addition, the exposure can be controlled manually so that the PC is not needed on site - you use switches in the way you might use a cable release on a normal camera. Jon. From: "rjsdotorg" Date: Tue Feb 19, 2002 8:44 am Subject: Re: ExView, more results Reading the CCD output to a 16bit A/D might be an expensive option unless you want to slow it down a lot - the data is clocked out at a pretty high rate. 640*480X25fps = nearly 8M samples per second. If you slow it down, you lose the real-time advantage of video. You could have it switchable I guess. That would be the advantage of having the existing camera circuitry, it could still be used to focus and find. 8bits would probably give better quality than is currently achievable using a capture card, since there would be much less opportunity for noise to creep in before digitisation. I suspect that the NTSC signal reliably provides more than 256 levels of grey, probalbly more like 10 bit. The noise reducion is the key, we are doing analog-RF-digital conversion; I read that most of the heat is from the driver chip (24ma) and the processor(45-60ma!) and not the CCD (only 6ma), so cooling the CCD separately is 'easier', although the hotter chips would inject less noise is they ran cooler. Given the noise levels in the cheap CCDs used in these cameras, I would think going to 12 or 16 bits might be overkill anyway. If you want higher bit depth, take more images and sum them. Although inexpensive, the Exview compares well at low temps to much more expensive, older designs. A relative who is a CCD chip-designer figured that the accessory electronics are the definite limiters of quality here. Check out the specs on their CCDs: CCD Specs They sell for $20k to $200K! Per chip! But if you soldered 'em to these circuit boards, they wouldn't look much better. Since the limit to summing is roughly where: ave signal < (charge-well noise + readout noise)*bit depth no matter how many dark frames or sums done, your well-cooled chip is probably running up against readout noise now, so by-passing the processor to direct A/D is probably the next step, and likely to reveal orders of magnitude deeper objects. Of course, I'm still waiting for one clear San Diego night to even try mine on the Celestron. And, unfortunately, the relative is a physicist and not an EE. BTW, I found a Videum AV card (does both PAL and NTSC) at the swap meet for $5. If anyone else is interested, I'll look for more similar next time. (I was using the ATI A-I-W Pro on this machine.) Just my 2 penn'orth, It's worth a lot more than that! Ray From: "rjsdotorg" Date: Tue Feb 19, 2002 8:51 am Subject: Re: ExView, more results --- In QCUIAG@y..., Steven Jindra wrote: Jon How much data would be lost if the live video is recorded to a Hi-8 camcorder or in my case a Digital-8 then later reviewed frame at a time to capture the high resolution, long exposure frame(s) to computer through a capture card? My new Hi-8 is much clearer than 8mm or VHS, and the digital-8 should not lose (almost) any if you use the firewire output type. The only loss would probably be in the conversion/transmission of NTSC signal. Ray From: Jon Grove Date: Tue Feb 19, 2002 9:00 am Subject: RE: [QCUIAG] Re: ExView, more results In that case it would definitely be an option. I found manual control of exposures quite satisfying and immediate. A disadvantage is that it's hard to focus since the snapshot is there and gone in a 25th (or 30th) of a second. But if you can focus first on a bright star you should be OK. Of course you still need a capture device in the end, but you don't need to take your PC out to the scope. Jon. From: "wlbehrens" Date: Tue Feb 19, 2002 10:06 am Subject: Re: ExView, future schemes Already working on this Jon and moving the data via firewire. This is the next step in building this beast :). Can't wait for your schematic......I am most interested in the rumored mechanical parts (relay) in your design. William From: "Dr. S.J. Wainwright" Date: Tue Feb 19, 2002 10:15 am Subject: Re: [QCUIAG] Re: ExView, future schemes Jon, It is easy to clarify this one. An established home-built like the Cookbook and the Audine/Genesis would be outside our scope as they are already worked out, established in the marketplace and thoroughly conventional in that sense. Anything that arises de novo from our members, be it a camera mod or a completely new design is definitely part of our remit. Good luck with all innovation. Clear skies Steve Dr. S.J. Wainwright www.digital-astronomy.org.uk FAX: UK: 0870 1641692 International : +44 870 1641692 ----- Original Message ----- From: "Jon Grove" To: Sent: Tuesday, February 19, 2002 5:14 PM Subject: RE: [QCUIAG] Re: ExView, future schemes I'm not sure where the line is drawn, beyond which a modified camera falls outside the realms of this group. My feeling is that by the time you've added A/D, memory, D/A, slow clock, etc. the camera is no longer a modified CCTV camera, but has become a home built astronomical camera which happens to have some bits from a board cam. Having said that, I'd still like to build one! My scheme would be to put a D/A immediately on the CCD output and have it write its output to a RAM (512K 'words' (8,10,12 bits?) would be enough). When not being written to, the RAM would be read and its output fed to a D/A, and thence to where the CCD output would have gone. In other words there would be a framestore, so that on capturing a frame we can set it to keep transmitting that data until the next exposure comes along. This is what the Mintron seems to do so effectively. In addition it would be useful to have a slower readout of the RAM to enable a noiseless download of the data via a parallel or serial port instead of capturing via a capture card. In order to use the existing timing generators to provide the video output it'd be easiest to not slow the clock down, which is why I suggested an 8bit ADC. I couldn't find an affordable 10 or 12bit one that would work at around 10MHz. If there is one, let me know! Cheers, Jon. -----Original Message----- From: rjsdotorg [mailto:yahoo@r...] Sent: 19 February 2002 16:45 To: QCUIAG@yahoogroups.com Subject: [QCUIAG] Re: ExView, more results Reading the CCD output to a 16bit A/D might be an expensive option unless you want to slow it down a lot - the data is clocked out at a pretty high rate. 640*480X25fps = nearly 8M samples per second. If you slow it down, you lose the real-time advantage of video. You could have it switchable I guess. That would be the advantage of having the existing camera circuitry, it could still be used to focus and find. 8 bits would probably give better quality than is currently achievable using a capture card, since there would be much less opportunity for noise to creep in before digitisation. I suspect that the NTSC signal reliably provides more than 256 levels of grey, probalbly more like 10 bit. The noise reducion is the key, we are doing analog-RF-digital conversion; I read that most of the heat is from the driver chip (24ma) and the processor(45-60ma!) and not the CCD (only 6ma), so cooling the CCD separately is 'easier', although the hotter chips would inject less noise is they ran cooler. Given the noise levels in the cheap CCDs used in these cameras, I would think going to 12 or 16 bits might be overkill anyway. If you want higher bit depth, take more images and sum them. Although inexpensive, the Exview compares well at low temps to much more expensive, older designs. A relative who is a CCD chip-designer figured that the accessory electronics are the definite limiters of quality here. Check out the specs on their CCDs: Specs They sell for $20k to $200K! Per chip! But if you soldered 'em to these circuit boards, they wouldn't look much better. Since the limit to summing is roughly where: ave signal < (charge-well noise + readout noise)*bit depth no matter how many dark frames or sums done, your well-cooled chip is probably running up against readout noise now, so by-passing the processor to direct A/D is probably the next step, and likely to reveal orders of magnitude deeper objects. Of course, I'm still waiting for one clear San Diego night to even try mine on the Celestron. And, unfortunately, the relative is a physicist and not an EE. BTW, I found a Videum AV card (does both PAL and NTSC) at the swap meet for $5. If anyone else is interested, I'll look for more similar next time. (I was using the ATI A-I-W Pro on this machine.) Just my 2 penn'orth, It's worth a lot more than that! Ray From: Jon Grove Date: Tue Feb 19, 2002 10:16 am Subject: RE: [QCUIAG] Re: ExView, future schemes Not just a rumour! There are three of them, and they make a lovely clicking sound to let you know the thing's working ok. I tried various alternatives but the relays were the only option that worked. In case anyone wants to get the bits for the mod, here is a list which I think covers everything: 3 x 12V subminiature SPCO relay (Maplin UG42V) 1 x SN7406 hex O/C inverter (Maplin QX75S) 4 x small 1N4148 diode 1 x 10K resistor 1 x 0.1microF capacitor (polyester I think) Piece of stripboard about 2"x2", maybe a bit bigger to be on the safe side. Some ribbon cable for connecting the daughterboard to the camera PCB. At least 7-way, possibly 10-way. Connector of your choice for connecting to the PC. Needs to be able to carry the following signals: +12V power to camera Ground Video output 4 data lines from parallel port I also found it convenient to use a plug/socket to connect the daughterboard to the camera. I used a DIL IC socket cut in half lengthwise which made both a plug and a socket, but not all socket types will be useable. A 10-way IDC ribbon cable connector ought to be enough, or you can get away with 7-way if you use the 3-way power/video/ground connector on the camera PCB as well. I'm sorry to be taking so long publishing the design - it will happen any day now, honest! Jon. F From: "Steven S. Herrick" Date: Tue Feb 19, 2002 10:51 am Subject: RE: [QCUIAG] Re: ExView, future schemes Jon - I don't quite understand what it is that you're doing but I can't recommend that you turn your stored digital data back into analog for (eventual) redigitization unless you really have to. You might look at the Sony CXD2311 for an ADC ... 10 bits at 20MHZ but SMT mounting only for $10.15. Steve From: Jon Grove Date: Wed Feb 27, 2002 1:42 am Subject: ExView dark/cooling test Last night I tried some experiments with the ExView camera to see how cooling it affected the dark current/noise characteristics. The results are attached - each swatch is a 100x100 section from the top left of the full 640x480 frame. I switched the camera on, uncooled, and took a sequence of three 4-minute exposures with the 'lens cap' on. (The front of the camera is a sawn-off film cannister, the lens cap is the lid from one but it fits onto the sawn-off end so isn't very positive). As you can see, the noise levels gradually increased as the camera warmed up, although the hot pixels remained largely constant. The background grey level is because the lens cap wasn't very light-tight. I then switched on the Peltier and took another three 4-minute exposures while the camera cooled down. as expected, the noise level decreased until it was better than the first uncooled frame. Also the hot pixels seemed to improve. I then realised that light was leaking in to the camera so I turned off the room lights and put a cloth over the camera, and took another 4-minute exposure. This looked pretty much like the previous one only darker, as expected. So I tried a 10 minute exposure and whilst there was some noise visible I was quite surprised at how dark the frame was. Finally I turned off the Peltier (which would allow the CCD to warm up, possibly to a higher temperature than it would normally reach because of heat conducted from the hot side of the Peltier) and took another 10 minute exposure. This was very noisy as you can see. So it appears that it is worth cooling these devices for long exposures - even if only by using ambient air. It's interesting to note that the noise has a very similar pattern in each of the images, which indicates that it might be largely treatable by darkframe subtraction. I may need to do some more experiments to determine how much of the 'noise' is random (thermal) and how much is fixed (dark current). Food for thought, anyway. Jon Grove. Attachment DarkTest.jpg Type: image/jpeg Size: 49k From: Jon Grove Date: Wed Feb 27, 2002 4:23 am Subject: RE: [QCUIAG] Re: ExView dark/cooling test Richard, I think you're right - you could probably go to about 5 mins uncooled and use postprocessing to reduce the effects of dark current. You'd have a higher noise level on top of the dark current, which you couldn't get rid of (other than by stacking/summing/averaging multiple exposures). But I need to do some more experiments to work out how much random noise there is in the warm vs cold exposures. Good point about the flatframe generator - I hadn't thought of it like that! Jon. From: "rjsdotorg" Date: Wed Feb 27, 2002 9:25 pm Subject: Re: ExView dark/cooling test I've been following the thread with interest as I'm now working/soldering on my 1004x. After re-reading CCD Astronomy (C. Buil, 1991 <-ancient!) and reflecting on it a bit, I have decided to use the dry ice method, and cool the entire board. His example of calculating thermal charge for a small CCD in a small 'can' casing is like this: mW element 200 CCD array chip 1430 thermal conduction via the very thin wire leads 152 thermal power absorbed by the chip at -20C (19mW/cm2) 280 loss by convection from the casing, with a mild vacuum (.01 Torr) in the housing 810 'loss' due to cold finger radiation absorbtion, with 15sq.cm total surface _____ 2.87 W The wire leads conduct heat to the chip very well, and that is with .5mm wire, so I plan to use thermal caulk ($20-8oz tube) to thinly pot the whole back to a copper box, filled with CO2 from Baskin Robbins ($1.35/lb). One pound requires 241 BTU to sublimate, about .07 kW/h, and should =~20 hours of observing. I also considered the 12v-10A peltier I already had, with glycol circulated through brine ice at -10 to -20C. By the author's experience, both methods acheive about -50C chip temps, but even though I have all of the parts, and I have used the cooler to cool a very small salt water fish tank, I think the CO2 will be simpler to use in the end. Also from Buil, thermal signal decreases by a factor of 2.5 for every 10C decrease, which translated to ~10sec @27C, 1min @0C, and 10min @ - 30C for over-saturation; it also increases max image sensativity by 1 magnitude. Since my scope is a C114HD, it could use the help. As for summing, it can only help above where the radiation from the object is greater than the thermal noise, which grows with the root of the average thermal signal. Frost was considered a real issue, and even dry gas was not reliable under -20C, unless the cold finger was significantly colder than the chip to attract all the water first. I plan to try my welding gasses, I would need to get an old coated 'lens protector' filter in a small size so that vacuum doesn't distort it. Thanks to all for the great pictures, plans and inspiration, Ray http://rjs.org/astro Jon's mention of bumping the image ~10 pixels between images also seems to be common to remove large-scale noise or artifacts. From: "ssb73q" Date: Wed Mar 6, 2002 4:39 pm Subject: First light test of unmod 1004X ExView HAD camera My RF Concepts 1004X (actually a 1004XA, $68USD total cost, camera + shipping) arrived at my location (NY state, USA) ~10 days after ordering from RF Concepts: http://www.rfconcepts.co.uk/board_camera_bw_low_light.htm This camera is the same model camera that Jon Grove has modified for long exposure see: http://www.geocities.com/jgroveuk/Astrophotography.html The camera delivered was the 1004XA model that has a daughter board on it for sound. The first thing I did was remove the daughter board. I then mounted it in a small gadget box and attached phono jack, power jack, and a Edmund Scientific C-mount adapter to connect the camera to my NexStar 8. Now to the testing. Visual images from the 1004X of M42 on my video monitor were excellent. I compared the image with my SuperCircuit PC164C camera (both use the Sony Super HAD ExView 1/3" CCD) and was somewhat surprised. The 1004X gave better visual definition, obvious were the three bright sites just above the trapezium which were washed out in the nebula of the PC164C. The nebula image from the 1004X was also brighter. The 1004X is one heck of a camera value at $68USD delivered. Already on order is another 1004X and microfan to do Jon's modification. Regards, Richard From: julian huertas Date: Thu Mar 7, 2002 9:51 am Subject: increase of sensibility in 1004X ? Hi all Some weeks ago, I received from RF concepts my 1004X Exview. The first thing that I made was to prove the camera and to check their high sensibility. But yesterday, manipulating the main board, they came unfastened two resistances accidentally, next to the CCD. The two resistances are marked with a red circle in the picture.Later on, so that the disaster was complete, I lost one of them, the one that has marked the figure "473". Did I get resistances of substitution of an old Walkman that component SMD also has, but was the problem to know the exact value of the resistance, 47 Kohms?, 4,7 Kohms? , 473 ohms? , so that I proved with all them. I am not very sure, but I believe that the sensibility of the camera increases considerably if he/she diminishes the value of this resistance. This is a question for Jon Grove and for all those that work in the modification of this camera: does somebody know which the function of this resistance is? , is it possible that it varies the sensibility of the camera modifying their value?. Thank you in advance Julian. From: "resunltd4u" Date: Thu Mar 7, 2002 11:08 am Subject: Re: increase of sensibility in 1004X ? Hi Juilian, The value of a 473 resistor is 47K ohms or 47000. The third digit is the multplier. The marking "471" would be 470 ohms for another example. Regards, Bill Domenico From: Jon Grove Date: Thu Mar 7, 2002 10:08 am Subject: RE: [QCUIAG] increase of sensibility in 1004X ? Hi Julian, I'm not sure exactly what these resistors do, but it's possible they are part of the output amplifier for the CCD, or maybe the automatic gain control system. If so, changing them may well make the image brighter but will also increase noise levels by the same amount. I don't think you'll be able to increase sensitivity without sacrificing something. However it's worth investigating as it may enable dimmer objects to be imaged when combined with off-chip integration. I think 473 is 4.7K Ohms, and 102 is 100 Ohms. Thanks for posting the results of your 'experiments'! Jon Grove.