IT-LIST Digest 46Topics covered in this issue include:  1) Re: Snappy -- Here's my $0.02	by Ron Neumeyer   2) Re: Snappy -- Here's my $0.02	by SolamereTG@aol.com  3) Re: Snappy -- Here's my $0.02	by "Institute for Diabetes Research, Munich"   4) Re: Snappy -- a few more cents	by cxhx@musica.mcgill.ca  5) Re: Snappy -- Here's another 20 mils	by Robert Smith   6) identifying objects	by ap2@basil.acs.bolton.ac.uk  7) Re:Identifying Objects	by Alan Hale   8) Re: Snappy -- Here's another 20 mils	by Bryan Crosby   9) Re: Snappy -- Here's my $0.02	by Jeff Stiegman ----------------------------------------------------------------------Date: Tue, 10 Sep 1996 15:56:18 -0700 (PDT)From: Ron Neumeyer To: it-list@sparky.uthscsa.eduSubject: Re: Snappy -- Here's my $0.02Message-ID: <199609102256.PAA17143@zorro.bctel.ca>I use a Snappy with great sucess.  Try feeding the problem cameras through aVCR or Monitor with video in/out then to the Snappy.  The longer the cablethe more interference you will encounter. Regards,Ron*********************At 04:48 PM 9/10/96 +0600, you wrote:>Regarding Herek Clack's post I would also like to add a few "Snappy">comments and ask for advice.>>I also finally gave in a few months ago and bought a Snappy.  I think it>could be really cool for those working with VCR, camcorder and other types>of systems which use what I think are called "F-type" audio/video cables.>However, in my lab we have CCD cameras which use BNC cables.  In order to>make Snappy work I had to get an adapter (only $5 but I wonder if this>affects the performance.)>>The Snappy then works fairly well on one of our cameras, an 8 bit B/W, but>there are often interference lines running across the image.  On our color>video camera (just less than 8 bit) the only way to get an image is to wheel>the computer over right next to the camera and use its short 6' cord>directly into the Snappy.  Even then the image is extremely noisy.  Add any>longer BNC cords and the noise is too great for Snappy to recognize the>signal.  Also I had a demo of a 10-bit color camera in my lab and tried to>"snap" an image but it came out with a "shadow" double effect (perhaps due>to the contortions of cables needed to hook to my computer on short notice?).>>Any advice on what I could do to improve the signal??  I am trying to>acquire light microscopic images of histological sections.  My computers are>not located in the same rooms as the microscopes (no space).  Has anyone>successfully used a high resolution (10-bit or better) color camera with>Snappy?  >>A second problem is with Windows NT.  Like a fool I chose NT over Windows 95>when I found I couldn't run IT on Win 3.1.  Now I can't get NT drivers for>many of my beloved peripherals (e.g. Snappy and Bernoulli drive).>Supposedly NT runs much more smoothly than Win 95, but who cares if you>can't get drivers or installation setups for your software and hardware!>(So far there is no love lost between me and NT).>>Any comment would be appreciated!  TIA,>>Karen>>>>>>>...>>Just wanted to add something to the list (and archives) regarding Snappy,>>since in the past people have asked about it but no one seemed to have>>had first hand knowledge (or no one responded to the full list with such).>>>>Last week I finally broke down and went out and bought a Snappy Video>>Snapshot (Play Technologies makes them and their web page at >>www.play.com has the retail locations) 'cause its description,>>capabilities, and price seemed to be too good to be true, compared to>>the seemingly endless list of caveats, warnings, potential pitfalls,>>drawbacks and limitations of existing framegrabber setups.  Something>>that's inexpensive, flexible and capable, and is designed for >>plug-n-play use is too good to be true, no?...>>>Let me just say that Snappy is wonderfully easy to use, generates>>images in a number of file formats which generally seem to be>>compatible with IT (i.e. you can "open" the files rather than >>"importing" them), and with resolution that's excellent>>to my untrained eye.  Even paused video images are captured>>with no fuss, no muss, with the quality depending on the VCR in >>use (I have a good VCR, so capturing paused images -- which I'd>>been told would be hard to do well with framegrabbers -- has been>>very successful).  BTW, I'm using a comsumer-grade Hi-8 camcorder....>>>Herek L. Clack>>Combustion Toxics Laboratory>>Department of Mechanical Engineering>>University of California, Berkeley>>>>>>Life Sciences Sector Lab		  Reply:  kszaruba@mmm.com>3M Company>3M Center 270-1S-01			  Phone:  612-737-2971>St. Paul, MN  55144-1000                    Fax:      736-1519>>These opinions are my own and may not represent those of 3M.>> >>------------------------------Date: Tue, 10 Sep 1996 23:51:46 -0400From: SolamereTG@aol.comTo: it-list@sparky.uthscsa.eduSubject: Re: Snappy -- Here's my $0.02Message-ID: <960910235146_305708475@emout15.mail.aol.com>Dear KarenThe BNC cable is a 75 ohm cable that if not terminated properly can generatereflections (viewed as noise). As someone has already pointed out, going thrua VCR or monitor may clear up the noise (by providing the proper 75 ohmtermination) for your camera. I would guess that if the Snappy had a  75 ohmtermination, the cable effects would not be seen. G. PeetersSolamere Technology Group------------------------------Date: Wed, 11 Sep 1996 11:19:11 +0200From: "Institute for Diabetes Research, Munich" To: it-list@sparky.uthscsa.edu, kszaruba@mmm.comSubject: Re: Snappy -- Here's my $0.02Message-ID: <9609110919.AA06342@sparcserver.lrz-muenchen.de>Hi. If your image is too noisy cause there are too much losses in the cable, youcould check two things: 1) What is the quality of the cable? If you use it in a "dirty" environment,maybe a high performance higly shielded cable could be required. Just toshield the high frequency signal coming out from computers and otherelectronical equipment. Also check the connectors, us as less of them aspossible2) Check if your camera has an output gain control. To get a highsignal/noise ratio, a high video output level should be achieved. If youhave an electronic technician, maybe there's a way to manipulate the outputgain directly on the circuit board. If nothing helps at all, maybe insertinga small amplifier into the camera *before* the cable could do it.Was probably carrying owls to athens;-)Wolfgang Schechinger*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*Institute for Diabetes Research       *Koelner Platz 1                       *80804 Munich                          *  Germany                               *Phone +49 (89) 30 79 31 0             *Fax   +49 (89) 30 81 73 3             *  email diabetes@lrz.uni-muenchen.de    **-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*------------------------------Date: Wed, 11 Sep 96 07:45:53 -0500From: cxhx@musica.mcgill.caTo: it-list@sparky.uthscsa.edu, cxhx@musica.mcgill.caSubject: Re: Snappy -- a few more centsMessage-ID: <9609111447.AA19157@sparky.uthscsa.edu>The image distortions and contortions can be related to power outlet groundingbetween camera and computer. The usual noise abatement techniques are in order.First try different power outlet combinations, then try insulating the 3rdpin(s) . If interference is due to lead pickup, the length and shielding ofthose can be modified too.Paul HerouxMcGill Medicine------------------------------Date: Wed, 11 Sep 1996 12:42:20 -0400From: Robert Smith To: it-list@sparky.uthscsa.eduSubject: Re: Snappy -- Here's another 20 milsMessage-ID: <1.5.4.32.19960911164220.00685830@curtech.com>At 11:21 PM 9/10/96 +0600, you wrote:>Dear Karen>The BNC cable is a 75 ohm cable that if not terminated properly can generate>reflections (viewed as noise). As someone has already pointed out, going thru>a VCR or monitor may clear up the noise (by providing the proper 75 ohm>termination) for your camera. I would guess that if the Snappy had a  75 ohm>termination, the cable effects would not be seen. >G. Peeters>Solamere Technology Group>>Karen:  M. Peeters has a good point about terminating the cable.  If people aren'tfamiliar with how to do this it's quite simple.  You can buy a 75 ohmtermination for slightly more than $0.02, but not much.  Place this on oneside of a T-adaptor with the imput cable and Snappy connected to the other 2sides.  The closer you are to the Snappy, the better; optimally the T willplug right into it, possibly with a gender-changing adaptor.  I notice theSnappy has a "thru" connection;  probably just putting the 75 ohms acrossthat will do it just fine. If you're good with a soldering iron, you cansolder a 75 ohm carbon resistor onto a connector and plug it in.    Oneproblem; if the video source is 75 ohm terminated -- as it should be -- thisresistor forms a voltage divider, halving the imput to the Snappy.  Videodevices are supposed to compensate, but the S/N ratio may suffer.  In thiscase, you would need a video amplifier.  Has anyone tried a consumer-typeamp (ie Radio Shack)?  They certainly have the hi-frequency bandwidth, but Idon't know if they go low enough.  If so, this would be very cheap.  If youwant to do it yourself, Maxim semiconductors has a very nice line of videobuffers.  Some of these double the signal w/o external parts to compensatefor the voltage division effect I mentioned above.  But you'll have to begood with a soldering iron, these are just raw ICs.Good Luck             .  Robert A. Smith, Ph.D.  _____    .    Vision Systems' Analyst |     |<.      Current Technology, Inc. |_____|   .    (603) 868-2270     ^       .  ras@curtech.com    / \   /   \------------------------------Date: Wed, 11 Sep 1996 18:46:07 +0100From: ap2@basil.acs.bolton.ac.ukTo: IT-LIST@sparky.uthscsa.eduSubject: identifying objectsMessage-ID: <009A8399.F4A4D00B.107@basil.acs.bolton.ac.uk>Date sent:  11-SEP-1996 18:40:53 Hi, I am new to using image tool, but it seems to be just what I need. I am trying to use it to calculate areas in scanned images of leaves.  When I use the Find objects command, certain leaves are not recognised as objects.  Why would this be?  I have tried tweaking the thresholding and hump size options but to no effect.Alicia Prowseap2@bolton.ac.uk------------------------------Date: Thu, 12 Sep 1996 09:53:45 +0100From: Alan Hale To: it-list@sparky.uthscsa.eduSubject: Re:Identifying ObjectsMessage-ID: <199609120853.JAA01146@captain.celtic.co.uk>Hi, Alicia. This is what I do (with images of lichens), though there may be a better way:1 Open your scanned image in Paint Shop Pro or similar.2 Using the "lassoo" tool trace the outlines of your objects to select them (hold the Shift key to trace multiple outlines.3 With the background colour set to black, Cut the objects from the image - then Cut again.4 Change the background colour to white, and Paste as New Image.You should now have your objects in black on a white background. This new image should be saved, then opened in Image Tool, where the Find Objects tool  should  have no problem recognisingand marking  the objects.I use a scale in the original image to calibrate Image Tool, then ObjectAnalysis to find Areas, Perimeters etc.Works fine for me, though a bit long-winded. maybe someone has a better way? I'd like to hear. Let me know if I can help more.Alan Haleno trouble > > Hi, I am new to using image tool, but it seems to be just what I need. > I am trying to use it to calculate areas in scanned images of > leaves.  When I use the Find objects command, certain leaves are not > recognised as objects.  Why would this be?  I have tried tweaking the > thresholding and hump size options but to no effect.> > Alicia Prowse> ap2@bolton.ac.uk> > Alan D. Hale     /----------||---\\------||------||---> Not the Skipper    //\\        ||    \\     ||      ||   //  \\       ||     \\    ||      ||  //----\\---   ||     //    ||------||---> //------\\---\ ||    //   /-||------||--->//        \\   \||---//---/  ||      ||------------------------------Date: Thu, 12 Sep 1996 10:05:55 +0100 (BST)From: Bryan Crosby To: it-list@sparky.uthscsa.eduSubject: Re: Snappy -- Here's another 20 milsMessage-ID: On Wed, 11 Sep 1996, Robert Smith wrote:> At 11:21 PM 9/10/96 +0600, you wrote:> >Dear Karen> >The BNC cable is a 75 ohm cable that if not terminated properly can generate..> >G. Peeters> >Solamere Technology Group....>   M. Peeters has a good point about terminating the cable.  If people aren't> familiar with how to do this it's quite simple.  You can buy a 75 ohm> termination for slightly more than $0.02, but not much.  Place this on one> side of a T-adaptor with the imput cable and Snappy connected to the other 2> sides.  The closer you are to the Snappy, the better; optimally the T will> plug right into it, possibly with a gender-changing adaptor.  I notice the> Snappy has a "thru" connection;  probably just putting the 75 ohms across> that will do it just fine. If you're good with a soldering iron, you can> solder a 75 ohm carbon resistor onto a connector and plug it in.    One.......Or I wonder if a quick chat to a radio technician would not produce a 'passive' impedance matcher. If impedance mismatch is the problem thenacouple of pounds worth of variable capacitors, pence worth of fixedcapaciators, a coil, a box, 2x 75 ohm BNC connectors a drawing a a pinetwork and no amplifier, less signal reduction(some loss in coil andcheap capacitors is expected, characteristic small time delay and voilaanother possible solution, vizsnappy==75 ohmBNC -->balancing Pi network--<75 ohmBNC==sourceof course maybe your snappy is not happysincerely yoursBryan Crosby------------------------------Date: Thu, 12 Sep 1996 09:00:52 -0400From: Jeff Stiegman To: it-list@sparky.uthscsa.eduCc: bpc@msen.comSubject: Re: Snappy -- Here's my $0.02Message-ID: <32380984.4C94@mail.msen.com>Karen S. Zaruba wrote:> > Regarding Herek Clack's post I would also like to add a few "Snappy"> comments and ask for advice.> > I also finally gave in a few months ago and bought a Snappy.  I think it> could be really cool for those working with VCR, camcorder and other types> of systems which use what I think are called "F-type" audio/video cables.> However, in my lab we have CCD cameras which use BNC cables.  In order to> make Snappy work I had to get an adapter (only $5 but I wonder if this> affects the performance.)> > The Snappy then works fairly well on one of our cameras, an 8 bit B/W, but> there are often interference lines running across the image.  On our color> video camera (just less than 8 bit) the only way to get an image is to wheel> the computer over right next to the camera and use its short 6' cord> directly into the Snappy.  Even then the image is extremely noisy.  Add any> longer BNC cords and the noise is too great for Snappy to recognize the> signal.  Also I had a demo of a 10-bit color camera in my lab and tried to> "snap" an image but it came out with a "shadow" double effect (perhaps due> to the contortions of cables needed to hook to my computer on short notice?).> > Any advice on what I could do to improve the signal??  I am trying to> acquire light microscopic images of histological sections.  My computers are> not located in the same rooms as the microscopes (no space).  Has anyone> successfully used a high resolution (10-bit or better) color camera with> Snappy?> > A second problem is with Windows NT.  Like a fool I chose NT over Windows 95> when I found I couldn't run IT on Win 3.1.  Now I can't get NT drivers for> many of my beloved peripherals (e.g. Snappy and Bernoulli drive).> Supposedly NT runs much more smoothly than Win 95, but who cares if you> can't get drivers or installation setups for your software and hardware!> (So far there is no love lost between me and NT).> > Any comment would be appreciated!  TIA,> > Karen> > >> >...> >Just wanted to add something to the list (and archives) regarding Snappy,> >since in the past people have asked about it but no one seemed to have> >had first hand knowledge (or no one responded to the full list with such).> >> >Last week I finally broke down and went out and bought a Snappy Video> >Snapshot (Play Technologies makes them and their web page at> >www.play.com has the retail locations) 'cause its description,> >capabilities, and price seemed to be too good to be true, compared to> >the seemingly endless list of caveats, warnings, potential pitfalls,> >drawbacks and limitations of existing framegrabber setups.  Something> >that's inexpensive, flexible and capable, and is designed for> >plug-n-play use is too good to be true, no?...> > >Let me just say that Snappy is wonderfully easy to use, generates> >images in a number of file formats which generally seem to be> >compatible with IT (i.e. you can "open" the files rather than> >"importing" them), and with resolution that's excellent> >to my untrained eye.  Even paused video images are captured> >with no fuss, no muss, with the quality depending on the VCR in> >use (I have a good VCR, so capturing paused images -- which I'd> >been told would be hard to do well with framegrabbers -- has been> >very successful).  BTW, I'm using a comsumer-grade Hi-8 camcorder....> > >Herek L. Clack> >Combustion Toxics Laboratory> >Department of Mechanical Engineering> >University of California, Berkeley> >> >> > Life Sciences Sector Lab                  Reply:  kszaruba@mmm.com> 3M Company> 3M Center 270-1S-01                       Phone:  612-737-2971> St. Paul, MN  55144-1000                    Fax:      736-1519> > These opinions are my own and may not represent those of 3M.> > Hi Karen -My company sells digital imaging systems and we run into the noise problems you described fairly often.  Let me give you my 2 cents worth.1.  The Snappy device is primarily designed for Y/C or S color video input.  These are composite video signals which take the color components from the CCD sensor (yellow/ cyan, or Red/Green/Blue) and approximate them using a single signal.  Your eye can't tell the difference but it doesn't work so well when you try to view a gray shade (B&W) signal.  You'll get what's called 'aliasing' in the final image, it shows up as a fine cross-hatch pattern in the image.  If Snappy has an anti-aliasing filter (in software or hardware) use it. It will remove this pattern.2.  Since you are using a Y/C signal, don't expect to get great quantitative information out of your image.  The only way to do that is with a three channel Red/Green/Blue (RGB) signal.  Even then, if the camera is a single chip CCD, the pixels in the array are masked as roughly 70% green, 20% red and 5% blue.  Put a blue filter in front of the image and you'll general see very poor resolution.3.  I believe Snappy device has a 75ohm input termination.  You don't need to terminate the cable again.  If you do you will load down the signal, the image will become dark.  However, if there actually are termination problems, or you are using a video cable with the wrong impediance, the 'noise' will look like small ripples coming off hard edges in the image.  This is due to the reflections in the cable as others have mentioned.  The best thing to do is to refer to the User's manual for you camera and use the cable it suggests.  It is probably 75 ohm (designated as RG59 or RG59/U), although I have seen 50 ohm used as well.4.  One of the fellows suggested power outlet grounding as a problem source, this is a good point.  Computer's in general are horrible sources of noise, particularly cheap ones.  I have seen huge differences in the noise put BACK onto the AC lines.  This is usually easy to identify.  Almost all noise generated by the switching supplies in computers will show up on your video monitor as regularly spaced diagonal lines.  These may be many fine lines running at a small angle or just a few lines running at a large angle.  The only way to eliminate them is to ensure that the computer power supply and your camera have a common ground.  That's not easily done, you might get lucky and a good earth ground will do the trick.  However, more often it's the DC ground for both the computer and camera that need to be tied together.  If you're not familiar with how to do this, don't try it.  Have a tech do it for you.Sorry for the long dissertation.  Hope it helps.Jeff------------------------------End of IT-LIST Digest 46