Skip to content

Weekly InfinaDyne Chat

November 6, 2015

This week is a little different with a video showing a demonstration of the Vindex product.

If you haven’t seen Vindex before, you might want to take a look at the web page for it.  It is a product that is designed to make life much simpler for anyone that needs to analyze or examine video as part of their job.  Most of the focus recently has been on processing surveillance video, usually from dedicated DVR systems, but Vindex is designed to work with nearly any sort of digital video.

Version 2 of Vindex was a significant improvement in that it added video enhancement.  Version 2.1 then added a new “frame reader” to Vindex so that it is no longer reliant on installing any sort of codec pack on Windows.  This demonstration video focuses on the video enhancement in sort of a trivial way – just increasing the brightness of a surveillance video so that we can see the subject’s face better.

There are certainly other ways to do this, but I think you will see from this demonstration that playing a video, grabbing a screen capture and then enhancing that with something like Photoshop is a bit more complicated and probably more time-consuming.  Also, while there are many very nice video tools that you can purchase, Vindex is priced to be within reach of every organization that is using digital video in any manner.

If you would like to try out Vindex for yourself, click here to go to the Vindex page and download the trial version.  Questions? send them to sales [@] infinadyne [.] com,

 

Advertisements

Weekly Chat, 23 October 2015

October 23, 2015

Here is another weekly chat, this time discussing some of the new features in CD/DVD Inspector version 5.0.  One of the big things with this version is that the even for simple tasks it is often the case that the same content is being retrieved from the disc multiple times.  If you bring up the new Image Gallery and then produce a report with thumbnails every single picture on the disc will have been read at least three times.  When there are thousands of pictures on a disc, this can be quite time-consuming.

Previously, there was a argument in favor of always reading from the disc.  Partly, this was somewhat outdated but still relevant for many, this argument was disk space.  A simple DVD that is full of data is going to take 4.35GB of hard drive space to hold an image and it has only been in the last few years that terabyte and multi-terabyte hard drives for computers were common.

Also, there was the argument that if all you are doing is producing a report from a disc creating an image file is of limited benefit if you then turn around and produce the report from the image file.  This is still true in some sense, although with the assumption of greater processing speed, faster hard drives and nearly unlimited hard disk space, it is more reasonable to create an image file from every disc as a first step.

When used interactively, this is very simple.  You point at a physical CD/DVD/Blu-ray drive and select it.  Immediately you are prompted for where to save an image file.  You can select between the “InfinaDyne” .IDIF/.DMEM format or an ISO+CUE format with the caveat that file system information is stored in the InfinaDyne format which eliminates re-discovery if the image file is opened again later.

For a robotic system the same option is utilized to control the processing of discs and image files, only for a robotic system the image file is stored in the predesignated output folder.  The format of this image file can be selected with the Robotic Loader preferences.  If no image file is selected, then the InfinaDyne image file format is used and the image file is deleted when processing of the disc ends.

The benefit of creating an image file as a first step of processing a disc is only really apparent when some of the newer features are used, but even if all you are doing is producing a report with a robotic system there maybe some benefit to running the report off an image file.

The option that controls this is called “Always collect image file” and it is on the general Options tab for Preferences.  It will be defaulted to checked with version 5.0 so if you do not want this behavior you will need to uncheck it.

If you are interested in something specific being discussed, please let me know at sales@infinadyne.com.

Weekly Chat

October 16, 2015

Here we have another weekly chat from InfinaDyne.  This time I’m talking about the DB Freedom product.

Things are progressing with CD/DVD Inspector 5.0 and there will be some major updates to FlashRetriever Forensic Edition.

Weekly chat

October 8, 2015

Here is the second in a series of video chats, this one talking about new features in CD/DVD Inspector version 5 and a mention of the Vindex product and what it can do for you if you are called upon to analyze digital video.

More coming next week!

Weekly Chat

October 3, 2015

I am going to try to post a video each week that is of general interest to InfinaDyne software users.  These will be initially hosted on YouTube, simply out of convenience.

Take a look at the first one:

CD/DVD Inspector 5 is something that is coming along and will be introducing a number of new features into the product.  There will be more features described in future video chats.

A UPS in a Virtual World

August 11, 2015

41zXOFArsvL[1]A while back we “converted” or “upgraded” from a number of separate machines to one big server machine running Xen.  There were a number of advantages to this approach for both capacity and management, as well as getting some increased horsepower for operations.  The one thing that didn’t make the transition easily was the way the UPS was being handled.

Like just about everyone, all the server machines are on a fairly large UPS and have been since the beginning.  The problem has always been handling notification between the machine the UPS is connected to and the rest of the machines.  There are a number of somewhat OK solutions for this, some pretty costly and some free.  Unfortunately the common theme for all of these is they don’t work all that well straight out of the box.

Recently, we upgraded the NAS from a Thecus to a Synology box and increased the capacity significantly.  The Thecus was getting a bit old and the software support for it has not been what it could have been – meaning existing customers more or less got forgotten by the company.  Synology has a far better software offering and is more compatible with the way things work here.  So we moved over to it.  As part of this, we needed to figure out a UPS sharing strategy that worked with the Synology box.

It turns out that there is one and only one UPS sharing solution that works with Synology – NUT or Network UPS Tools.  It is an open-source Linux/Unix tool originally but has been ported to both Mac OS and Windows in a sort-of satisfactory manner.  With this in mind, the implementation of NUT on all of the various servers (virtual and physical) connected to the “big” UPS was started.  It should be said that a “big” UPS these days is a lot smaller than you might think, if you aren’t buying UPSs for a company.  The days when a server could be expected to need a huge power supply for lots of hot hard drives is over.  Our server running Xen uses all 2.5 inch drives and even with dual redundant power supplies is very comfortably served with a 1500 VA UPS.

Of course, the first thing right out of the box that was discovered was that the documentation for Synology is great for sharing a UPS between multiple Synology boxes but utterly absent for sharing a UPS with anything else.  Synology uses a Linux base for their software and in keeping with the whole open-source idea the documentation was an afterthought.  NUT being open-source as well suffers from the same problem: nobody is getting paid to produce documentation and it just isn’t as sexy as writing code.

It seemed like the simplest alternative was to first plug the UPS into the Synology box and then figure out how to get the other “clients” to communicate with it.  It puts the onus of managing the UPS hardware connection onto Synology and there is no shortage of things saying that works.

So, with a number of Windows machines as “clients” for the UPS sharing, the next step was to find a Windows port of NUT that was complete enough to act as a client for the Linux Synology implementation.  What I managed to find was NUT-Installer-2.6.5-6.msi.  This was on the SourceForge web site for NUT and it is sort of complete.  It does include some dependencies that are installed with it, but somehow the builder of this package seems to have assumed that everyone has OpenSSL installed on their Windows machine.  It also makes some assumptions about it being OK to split up executable files into “bin” and “sbin” but not copy the DLL files in “bin” over to “sbin”.

What I did was find an OpenSSL .ZIP file and extract the necessary DLL files for NUT to work.  These files have to go in both the “bin” and “sbin” folders.  Another technique would be to properly install OpenSSL with these DLL files in something like the Windows\System32 folder – this didn’t seem to be necessary or all that desirable.

After finding that critical programs wouldn’t start properly from the “sbin” folder because of missing DLLs, all the DLL files from the “bin” folder were copied over to the “sbin” folder.  If you aren’t terribly adept at using Event Viewer you might not catch this problem right off and it will take considerable work to figure out what is going on.  I recommend just copying the files over and leaving it at that.  Another solution might be to move all the “sbin” stuff to the “bin” folder but that seemed like it might involve more severe configuration changes and might even require a recompile – which I did not want to do.

As a note aside, I didn’t want to recompile NUT for Windows because it is uses GCC and MSYS (a derivation or modern version of MinGW) and it would require installing a whole new toolset in order to compile it.  Sure it might be the “right” way, but it is also rather time-consuming so I really wanted a solution that would work from binary, pre-built files.

After figuring out how to tailor the configuration files for NUT on Windows to get it to work as a “slave” to the Synology box, I had three tailored configuration files out of six.  The others simply aren’t used.  There are a few key points when connecting a client to the Synology implementation:

  • The Synology UPS is called “ups”.  This doesn’t appear to be documented anywhere officially.  I did find a posting that referenced this but it is a fairly obscure and extremely necessary fact.
  • The user and password for connecting to the Synology UPS service is “monuser” and “secret”.  Again, no official source for this and I had to find it by digging through files on the Synology box.  Turns out I found this elsewhere in posts but buried pretty deeply.
  • There is a reference to a file in the UPSMON.CONF file and this must be changed from the default for Windows – the Windows distribution comes with the default set to something that will not work on Windows.

The last thing to do to get a client up and running is to start the “Network UPS Tools” service.  This is probably easiest done through the “Services” applet on Windows but I suppose one could use the command:

net start “Network UPS Tools”

with the caveat that the quotes are very important.

You will note that there isn’t any sort of indication externally that the NUT client is running.  It is hidden away as a service without any sort of user interface and no user-mode program that will say it is running properly.  There is a separate NUT client that is available called “winnut15.zip” out there that is actually a Windows user interface client.  All it does is display the status of a NUT “master”, so it isn’t useful for getting your Windows box to shut down in an orderly fashion when the UPS is running out of battery power.  But this program can be helpful in figuring out that the “master” is running and the UPS is doing what it is supposed to do.

The Synology box does not provide any indication of what “slave” or “client” computers are connected to it and therefore isn’t capable of telling you anything about your configuration.  How do you know everything is working?  I am going to say the only answer is to try it – unplug the UPS – and see what happens.

Trying to do this with other software, such as the APC stuff, wouldn’t have worked with any NAS box and it is not a nice installation on Linux in any respect.  Other open-source solutions exist, such as apcupsd, but they don’t get along with Synology.  I will say that the Synology implementation here seems to be very flexible in that while it is an APC UPS today there are a huge number of manufacturers that are supported by NUT and having the “master” being the Synology box isolates all of the clients from the details of the UPS hardware.

Science and Engineering

August 10, 2015

Previously, as in maybe 100+ years ago, “science” was important and the importance of it vastly exceeded the importance of engineering.  Today, I believe that trend has changed and the discipline of “engineering” is probably more important.  But, I am of the believe that engineering is at least the stepchild of science and you can’t ignore scientific discipline any more than you can ignore gravity.

Engineering, in layman’s terms, is the practice and discipline of making things that work, making things that aren’t working work, and the skills and abilities to be able to convert someone’s theory into practice.  This is certainly concerned with science but is more practically grounded.  In today’s world the distinction is important.

One of the key aspects of this is the motivation of the scientist is always “Why?” whereas in general the engineer isn’t really concerned with why or why not but just getting the thing working.  Out in the desert with a non-working car the scientist is going to be stuck on the point of trying to figure out where the water in the cooling system went and while he might have some interesting theories about where you could find water in the desert, the engineer is going to be focused a lot more firmly on both getting something into the cooling system and that it stays put.  The idea of using a children’s fruit drink isn’t going to occur to the scientist – its not water, remember – but to the engineer the sugar in the drink might just be the ticket to sealing up the leak.

McGyver is certainly not a scientist but he is the 1980s answer for engineering.

Over the weekend I made a big mistake; once again I watched the movie “The Core”.  This came out over 10 years ago and, as far as I know, has never been equaled in its mistakes, gotchas and distortions.  One of my least favorite lines is where one of the “scientists” in explaining to someone that all science is “best guess” and nobody really knows anything.  This might be a cute line for a post-modernistic movie tailored for an audience that believes science has given them nothing but Agent Orange and Tang but shows a disturbing lack of education.

Part of the annoyance of watching the movie is that for the entire 135 minutes there is about one silly thing every minute or so.  There are plenty of web sites that talk about the “bad science” of this movie, so many that there is no need to repeat the litany of wrong-headedness that pervades this movie.

The tie-in with this post is early on in the beginning of the movie we see pigeons crashing through big windows in tall buildings.  Now a “scientist” might be able to come up with a plausible theory how this is possible, but from an engineering perspective it is incredibly silly.  I did a very small amount of research on this last night and discovered the phrase that tempered glass is a pretty old technique and one that has been superceded by a number of other ways to make higher strength glass.  But even tempered glass in a 1/4 inch thick window glass form requires an impact of around 10,000 PSI to punch through it.  With a hammer, what you need is 10,000 pounds of force (better expressed in something like neutons, but I’m trying to stick with what I have) to break such a window – not impossible with a 1-pound hammer with about a 1/4 inch striking surface and a good swing at the window.  This is assuming the hammer is tilted somewhat when striking the window – if the head of the hammer was perfectly perpendicular to the glass the striking force would be distributed over a larger area and require even more powerful a swing.

OK, so what about a pigeon?  Your average pigeon is 9-13 ounces in weight, and the top of its little head might be at most 1 square inch in area after some initial mashing occurred.  So could a pigeon create a 10,000 PSI impact into a glass window?  I am NOT going to get into the full depth of the mathematics here to prove such a thing is impossible, but I am going to say that it would require the pigeon be flying over 100 MPH at the time of impact.  Highly doubtful.  I don’t think you could break a modern building window with a pigeon even if it was fired from a good-sized air cannon at over 100 MPH.  One of the key aspects of this is the little pigeon body has far less resistance to impact forces than the glass.  So even if you has sufficient acceleration of the pigeon, the pigeon would be pulped rather than actually breaking the glass.

Could you break the window with more pigeon impacts?  Say, 100 pigeons striking at the same time?  Dividing up the force required by 100 would certainly change things some, but still we are talking about 100 pounds of pigeon needing to develop far more than 10,000 pounds of force because it is spread over a much larger area.  The end result is 100 pigeons become a sticky mass of pulped pigeon and the window remains intact.

This commentary on pigeons can also be applied to the 1963 Alfred Hitchcock classic “The Birds”.  Although we are dealing with small window pane glass or the glass in a telephone booth, still the mismatch between bird structure and glass structure leads one to the inescapable conclusion that the birds simply do not have sufficient mass or cohesiveness to effect damage on virtually any size piece of glass.  At least for birds the size of pigeons or smaller.  Now, if you want to talk about bald eagles going up against a window the size of the bird itself, well, we are talking about a whole different sort of scale here.

Here is where the engineering discipline makes a significant difference.  A scientist is likely to focus on mass-on-mass impact and come out with a lot of equations showing that at 1027 MPH a pigeon can penetrate a large office building window.  The engineer is far more likely to note very early on that one reason pigeons do not fly over 30-40 MPH is because their wings would tear off.  Similarly, upon impact with any hard surface the less cohesive mass is going to come apart and distribute the impact energy over a much larger area.  There is little question that a pigeon is less cohesive than a big glass window.

So rather than expressing the problem as a purely mathematical one that has a solution – all it takes is a big blackboard – the engineer can get right to the heart of the question and say all the math is pointless because the pigeon comes apart.

This is the difference between the scientist and the engineer.

I would happily further this discussion with people that want to refute this or confirm it.