1. Source
• Magnetism is caused by moving charged electrical particles (Faraday, 1830s).  These particles can be the current of electrons through an electric wire, or the movement of charged particles (protons and electrons) within an atom.  These charged particles move much like planets in a solar system:
• nucleus spin around its own axis, causing a very weak magnetic field.
• electrons orbit around the nucleus, causing a weak magnetic field.
• electrons spin around their own axis at the speed of light, causing a significant magnetic field (Goudsmit and Kronig, 1925).
• Spinning electrons generate the bulk of the magnetism in an atom.
• Within each orbit, electrons with opposite spins pair together, resulting in no net magnetic field.
• Electrons in an orbit are filled up first by a + spin.  Once all the orbitals are filled with unpaired + spins, the orbitals are then filled with the – spin. (see Spin direction)
  :
2. Characteristics
• The electron configuration in an atom determines a the magnetic characteristics:
• In diamagnetic material, such as copper (Cu), all electrons are paired together.  There is no net magnetic field from the electrons.
• In paramagnetic material, such as magnesium (Mg), there are unpaired electrons.  The atoms align to an external magnetic field.  It becomes magnetized while the external field is present.
• In ferromagnetic material, such as iron (Fe), there are unpaired electrons also.  In this case, the atoms not only align to an external magnetic field, but also orient parallel to each other.  Thus, even when the applied field is removed, the electrons maintain in a parallel orientation.
• experiment: use a magnet to stroke a ferromagnetic bar, such as iron, several times in the same direction.  The magnetic force from the north pole of the magnet causes the unpaired electrons to align themselves.  The iron will stay magnetic.
• See also Electron spin, Theories of magnetism, How atoms work, and Goudsmit’s lecture.
• added challenge
• MIT lectures 8.02,
• NDT article,
• Relativity tells us that what looks like a pure magnetic ¯eld in one frame of reference looks like an electric ¯eld in another frame of reference.

Properties

1. Magnetics
• The magnetic field (B) is very strong at the poles and weakens as the distance to the poles increases.
• A magnetic field line (Faraday) is a theoretical line that loops through the north pole of a magnet, passing through surrounding space, enter the southing pole and going through the magnet back to the north pole.  A higher density of nearby field lines indicates a stronger magnetic field.  Field lines are a visual and conceptual aid only and are no more real than the contour lines (constant altitude) on a topographic map.
• simulation: move compass and bar magnet using Faraday’s Lab.
• experiment: visualize magnetic field using iron filings on paper/glass with bar magnet underneath.
  :
2. Two magnets
• Interaction
• Opposites attract.
• Same poles repel.
• The force between the two poles is directly proportional to the product of the pole strengths and inversely proportional to the square of the distance between the poles.
• A compass is tiny magnet balanced on a pivot.  It will rotate to point toward the opposite pole of a magnet.
• Experiment: suspend a magnet from a string.  It will align with the earth’s magnetic field, so that its S-pole points to the Earth’s N-pole.
  :
3. Induction
• A material that is attracted by a magnet becomes a magnet itself.
• As an iron nail is brought close to a bar magnet, some flux lines emanating from the north pole of the magnet pass through the iron nail in completing their magnetic path. Since magnetic lines of force travel inside a magnet from the south pole to the north pole, the nail will be magnetized in such a polarity that its south pole will be adjacent to the north pole of the bar magnet. There is now an attraction between the two magnets.
• experiment: move an iron nail to a bar magnet.  Touch another nail to the end of the first nail.  The magnetic field from the bar magnet will align the unpaired spins.  First the ones closest to the bar magnet.  This process can be repeated until the strength of the magnetic field weakens with the distance from the bar magnet.  As soon as the first nail is pulled away from the bar magnet, all the nails will fall. Each nail had become a temporary magnet, but once the magnetizing force was removed, the unpaired spins once again assumed a random distribution
• See also  …..

Electromagnetism

1. Wire
• A moving charge (current) through a wire produces a magnetic field (Ørsted 1819, Ampère 1820).
• The magnetic field lines form concentric circles around the wire.
• The strength of the magnetic field:  |B| = μ₀.|I| /2Πr, where μ₀ is the permeability of free space, I is the current, r is the distance from the wire.
• The direction of the field follows the right-hand rule convention for vectors.  In this case, the corkscrew rule, where the thumb points in the direction of the current, and the fingers point in the direction of the magnetic field lines.
• Experiment, paper with iron filings, wire perpendicular to this paper and few seconds of strong current (100 A) through the wire.
• See also the lecture Magnetic field and Lorentz Force, MIT.
  :
2. Wire in magnetic field experiences force
• A charge moving through a magnetic field experiences a force F_B^→ = q . (v^→ × B^→), where v is the speed of the charge, q is the value of the charge and B^→ is the magnetic field (Lorenz, 1892).
• The direction of F_B^→ is perpendicular to both the direction of the charge and the magnetic field.  The direction is the cross product of v^→ and B^→, and as such defined according to the right-hand rule for vectors.   With the thumb, index, and middle fingers at right angles to each other (with the index finger pointed straight), the first (index) finger represents the first vector in the product (v^→); the second (middle) finger represents the second vector (B^→); and the thumb represents the product (F_B^→).
• Animation, Lorenz force, Univ. of Florida.
• Experiment, current perpendicular to magnetic field.  Thin wire (50 cm) suspended in between strong magnets.  Connect wire to switch and two D-cell batteries.
• Experiment, wire in between two poles demo as part of the lecture Magnetic field and Lorentz Force, MIT.
• Experiment, Homopolar motor.  (see also Homopolar)
  :
3. Coil in magnetic field
• make one, http://science.howstuffworks.com/how-to-make-electromagnet.htm
4. Rotor in magnetic field
• Simple motor, using D-cell battery, magnet, 2 clothes pins, and coil. (video, video2)
5. Eddy currents.  Two equal masses slide down two identical aluminum tubes. Since one of the masses is magnetic it will induce eddy currents according to Lenz’ Law. This damps the motion of the mass causing the magnetic mass to take much longer to fall through the tube

other stuff

How voltage is produced: http://www.tpub.com/neets/book1/chapter1/1k.htm

• very strong at the poles and weakens as the distance to the poles increases;
• goes from the north pole of a magnet, pass through surrounding space, enter the south pole and go through the magnet back to the north pole, thus forming a closed loop.
• experiment: visualize magnetic field using iron filings on paper/glass with bar magnet underneath.

Homopolar motor, Wiki.

1. Book, “Science Explorer – From bacteria to plants, §1″ (Prentice Hall)
   • Grows
   • Uses energy
   • Responds to their surroundings
   • Reproduces

Where does life come from?

1. Life comes from other life, but where did the first life come from?
2. First a little bit of Chemistry
   • Book, “Science Explorer – Chemical Interactions Resources” (Foss Middle School)
     • A molecule is the smallest part of a substance that still has the same properties.  E.g. a water molecule is called H₂O.
     • Molecules are build out of atoms.   E.g. a water molecule consists out of two Hydrogen (H₂) and Oxygen (O₂) atoms.
     • Atoms are build out of protons, electrons and neutrons.  E.g.  a Hydrogen (H₂) atom consists of 1 proton and 1 electron.
   • Web page, “Atoms, Elements and Molecules Q&A” (Jefferson Lab)
   • Web page, “Atoms structure” (chem4kids.com)
3. Birth of our solar system
   • Web page, “Very brief history of the universe” (suite101.com)
   • Web page, “Origin of the chemical elements” (suite101.com)
   • Video, “Origins – Back to the beginning” (NOVA)
     • 13.75 billion years ago, there was a super explosion, from which a cloud of hydrogen (H) and helium (He) formed.
     • From the dust stars were born.  The stars were very big and inside the pressure and temperature extremely high.  H atoms fused together to form He atoms.  This released energy as heat and light.  When it ran out of H, they started to burn He instead.  This He atom fused together to form carbon (C), oxygen (O) and nitrogen (N).  These are the basic ingredients of life.
     • Once no more elements could be fused it stopped and the star exploded.
     • In the middle of a dust cloud, pressure and temperature rose and our Sun was born.
   • Video, “Origins – Earth is born” (NOVA)
     • Dust particles attached to each other and formed our planets.
     • 4.5 billion years ago, Earth was born.
     • Many comets fell into the earth.  These comets were big ice rocks from space that gave us water.
     • 4.4 – 4.1 billion years ago, the crust of the Earth cooled down and and water condensed.
   • Web page, “Abiogenesis” (Wikipedia)
   • Web page, “History of life” (Wikipedia)
   • Interview, “How did life begin” (NOVA)
   • Web site, “Exploring life’s origins” (Museum of science, Boston)
   • Web page, “Organic chemistry, biochemistry” (NY State Uni)
   • Web page, “Timeline of evolution” (Wikipedia)
   • Web page, “A hitchhiker’s guide to early earth” (Lawrence Livermore Lab)
   • Web page, “Miller-Urey experiment” (Wikipedia), “Miller/Urey experiment” (Duke Uni.)
   • Online book, “Amino acids, peptides and proteins” (Oxford Uni.)
   • Web animation, “Protein formation animation” (Montana State Uni.)
   • Web site, “Animating biology” (Carnegie Mellon Uni.)
   • Web page, “The RNA World and the origins of life” (US Nat’l Lib. of Medicine)
   • http://web.visionlearning.com/custom/chemistry/animations/CHE1.3-an-animations.shtml
   • Video, “Where did we come from?” (NOVA)
   • Video, “Origins – How life began” (NOVA)
     • All living things, are made from a small set of chemical elements: hydrogen (H), oxygen (O), carbon (C) and nitrogen (N) — four of the most common elements in the universe.
     • Life requires molecules that can catalyze reactions that lead, directly or indirectly, to the production of more molecules like themselves.  The following steps are might have generated live on Earth:
       1. Atoms of C, H, O and N can fill their outer electron shells by forming covalent bonds.  The resulting molecules include amino acids (Black for 2.16a), sugar alcohols such as ribose (2.11b) and glycerol (2.11c), fatty acids (2.13a), phosphoric acid (2.20b) and nitrogenous bases (2.22).  There are different theories of where these bonds could have formed:
          • in puddles (see “Where did we come from?“)
            • 2BD
            • Experiments show that the building blocks of RNA can assemble themselves naturally.  (John Sutherland)
          • near deep see volcanoes (see “Iron-sulfur world theory“)
            • Hot volcanic gasses (CO, NH₃, H₂S) in the deep ocean pass over a metal catalyst generate catalytic metallo-peptides. Further energy from sulfides of iron in redox reaction.  Energy available for the synthesis of organic molecules but also for the formation of oilomers and polymers.  Such systems may be able to evolve into autocatalytic sets of self-replicating, metabolically active entities that would predate the life forms known today.
            • Organisms collected at deep sea volcanoes are similar to some of the earliest organisms that existed on the Earth.
          • delivered by comets (see “Miller/Urey experiment“)
            • Lightning or UV-light from the sun can help in the creation of certain basic small molecules (monomers) of life, such as amino acids.  Amino acids are the basic building blocks of proteins and cells.
            • The energy from the impact of the comet fused amino acids together to form more complex molecules called peptides (< 50 amino acids linked together).
            • Experiments show that amino acids can assemble themselves naturally. (Miller–Urey)
       2. Covalent bonds between amino acids form long chains called proteins.  Proteins provide structure and control the rate of chemical reactions in cells.
       3. Covalent bonds between glycerol, two fatty acids chains and a phosphoric acid (H₃PO₄) creates a phospolipid.  The number of protons, electrons, energy level of the electrons and place in a molecule determine how much an atom attracts electrons  (see “electronegativity“).  An unequal relationship between atoms creates a polar covalent bond.  In phospholipids, the phosphate group is polar and can interact with water molecules, which are polar too, but the two long, uncharged fatty acid tails cannot.  As a result these molecules in water form lipid bilayers, a main component of the cell membrane.  The membrane provides a stable physical and chemical environment (see “Phospholipid membranes“)
       4. Covalent bonds between a nitrogenous base, ribose and phosphoric acid salt creates a nucleotide.  (see “Abiogenesis“)
          • Long chains of nucleotides form random (pre-)RNA molecules and result in self-replicating ribozymes (see “RNA world“).
          • Selection pressures for catalytic efficiency and diversity might have resulted in ribozymes which catalyse peptidyl transfer (hence formation of small proteins), since oligopeptides complex with RNA to form better catalysts. The first ribosome might have been created by such a process, resulting in more prevalent protein synthesis.
          • Synthesized proteins might then outcompete ribozymes in catalytic ability, and therefore become the dominant biopolymer, relegating nucleic acids to their modern use, predominantly as a carrier of genomic information.
     • Life evolves
       1. Selection pressures for catalytic efficiency and diversity result in ribozymes which catalyse peptidyl transfer (hence formation of small proteins), since oligopeptides complex with RNA to form better catalysts. Thus the first ribosome is born, and protein synthesis becomes more prevalent.
       2. Proteins outcompete ribozymes in catalytic ability, and therefore become the dominant biopolymer. Nucleic acids are restricted to predominantly genomic use.
     • aa
     • 3.9 – 2.5 billion years ago, simple cells “Prokaryotes” appear
       • They use CO₂ as a carbon source and get energy from oxidizing inorganic materials such as H₂S, FeO, H2 or NH₃.
       • Later, they evolve glycolysis, a set of chemical reactions that free the energy of organic molecules such as glucose and store it in the chemical bonds of ATP.  Glycolysis (and ATP) continue to be used in almost all organisms, unchanged, to this day.
     • 3.5 billion years ago, first archaea (3 RNA polymerases)
       • Organisms generate ATP by exploiting a proton gradient.
     • 2.7 billion years ago, photosynthetic cyanobacteria
       • They use water as a reducing agent, thereby producing oxygen as waste product.  The oxygen oxidizes dissolved iron in the oceans creating iron ore.
       • Snottites, single-cell bacteria that burn H₂S (?)
     • 2.5 billion years ago, cyanobacteria, oxygen appears
       • Microbes evolved a green pigment known as chlorophyll.  This allowed them to trap sunlight and use it to drive a chemical reaction that converts carbon dioxide (CO2) and water (H₂O) into food.  This “photosynthesis” enabled some bacteria to grow and reproduce almost without limit.  Life could spread, literally, over the entire planetary surface.  As a byproduct of photosynthesis, the ancient bacteria produced a waste gas: oxygen (O₂).
     • 1.85 billion years ago, first complex cells (Eukaryotes), primitive algae
     • 1.2 billion years ago, sexual reproduction, increasing rate of evolution
     • 1.2 billion years ago, multi-cellular organisms
       • The oxygen (O₂) level raised from 1% to 21% allowing multi-cellular life
     • 600 million years ago, first skeletons
     • 500 million years ago, first fish
     • 475 million years ago, first land plants
     • 400 million years ago, first insects and seeds
     • 360 million years ago, first amphibians
     • 300 million years ago, first reptiles
     • 200 million years ago, first mammals
     • 150 million years ago, first birds
     • 130 million years ago, first flowers
     • 2.5 million years ago, first genus homo
     • 2 million years ago, homo erectus
     • 0.2 million years ago, homo sapiens (similar to how we look today)
     • 0.006 million years ago, sumerians invent civilization
     • 0.002 million years ago, Roman Empire civilizes Europe
     • 0.000007 million years ago, you were born

Book, “Science Explorer – Cells and heredity” (Prentice Hall)

Questions

1. How are they not eaten?
2. What do they eat?
3. How do they get through the fall and winter?
4. How they reproduce?
5. How they live their lives?
6. How they get there food?
7. Differences  between plants and animals?
8. How they attack hungry predators?
9. How they digest their food?
10. How they are different from plants and animals?
11. Are they plants or animals?
12. How they look under a microscope?

Group your questions

1. Introduction (cells)
2. Plants
3. Animals
4. Protists

For each of the 3 groups answer

1. Species
   • Are they plants or animals?
   • How they look under a microscope?
   • Differences between plants and animals?
   • How they are different from plants and animals?
2. Feeding
   • What do they eat?
   • How do they get their food?
   • How do they digest their food?
   • How they live their lives?
3. Survival
   • How are they not eaten?
   • How they attack hungry predators?
   • How do they get through the fall and winter?
   • How do they reproduce?

(Alternate grouping)

1. Plants, protists and animals
2. Food chain
3. Life cycle
4. Under the microscope

Resources

Cells

• Web site “Classification of Living Things” (National Earth Science Teachers Association)
• Web site “Cell structure” (biology4kids.com)
• Web site “Biology course” (Uni. of California)
• Web site “Biology book” (Estrella Mountain CC)
• Web page “Eukayote” (simple Wikipedia)

Plants

• Web page What is the difference between plants and animals? (answers.com)
• Web page “Photosynthesis” and “Chloroplast” (biology4kids.com)
• Web page “Photosynthesis” (howstuffworks.com)
• Video “Photosynthesis” (howstuffworks.com)

Animals

• Web page “How cells obtain energy” (ebiomedia.com)
• Science project “Life is everywhere” (education.com)
• Web page “Cellular respiration” (wikipedia.com)
• Video “Cells, Tissues, Organs, and Systems” (nasa.gov)

Microorganisms

• Book “A world in a drop of water” (Alvin and Virginia Silverstein)
• Web page “Menu of microscopy and microscopes” (Microscopy UK)
• Web page “Photosynthesis, using algae wrapped in jelly balls” (Science and plants for schools)
• Teacher Notes “Pond water survey” (sciencespot.net)
• Web page “Biological specimens, fresh water” (micrographia.com)
• Picture “Algae sex-tape” (nationalgeographic.com)
• Web page “Paramecium” (101science.com)

This post is password protected. To view it please enter your password below:

Password:

Moving from CS3 to CS4/CS5

If you were one of the early adapters, you might have followed my guide Editing AVCHD 24p using Adobe Premiere Pro CS3.  If you made the jump to CS4 or CS5, you can not tweak your old projects so that they use the .mts files directly.  The following work flow did the trick for me:

1. Open the .prproj file using a text editor such as GNU Emacs, and search “.avi<” and replace it with “.mts”.  Once you are there, you may as well replace “<FrameBlend>true” with “<FrameBlend>false” to disable frame blending.
2. Move the .avi files (in the odd chance that you still have them) out of the way
3. Open the project in Premiere Pro.  It will automatically port it to the new version.  When it complains about the .mts being missing, just say “offline all”.
4. Once in Premiere, select all the .mts files and right-click.  Select “link media”, then point it to the .mts file.

Premiere Pro makes a fair attempt at removing the pull-down, but on high motion scenes, it misses the boat and doubles frames or leaves interlaced frames.  At first glance, playing with the the Field Options appears to help, but closer inspection of the rendered output will reveal doubled frames giving the video a stuttered look.  Those who shoot “24p wrapped in 60i” and expect a correct pulldown will need additional steps.

Oddly enough the companion product After Effects CS5 (AE) does a better job at removing the pull-down.  Maybe one day, the Premiere Pro developers will decide to reuse the .mts import functionality from After Effects.  Until then, you will need some additional legwork to extract the 24p from the 60i container.  For this, the following approaches come to mind:

1. Import the Premiere Project in After Effects.  Right-click each clip and have it Interpret and “Guess 3:2 pulldown”.  This can be automated by using a script such as ReverseTelicine.jsx.  The only downside is that AE only supports a subset of the Premiere Pro elements.  Titles and most transitions will be rendered as black video.
2. Import the clips in After effects and have it guess the 3:2 pulldown, create compositions and dynamic link those for use in Premiere Pro.  This will only work in the Premium version.
3. Use After Effects to convert the 24p-in-60i .mts files to 24p .mp4 files.  This can be achieved by importing the .mts files; select them all; and run ReverseTelicine.jsx to have it guess the cadence; select them all clips and add them to the Render Queue.  One would then import these 24p .mp4 files in Premiere Pro, or if you have an existing project using .mts files, you would replace the .mts assets with the equivalent .mp4 assets.  Note that AE still makes some mistakes, but beats PP.
4. Stick with the approach outlined in Editing AVCHD 24p using Adobe Premiere Pro CS3.  It takes a lot if disk space, but it is fast and accurate.  This might feel as two steps forward and one step back, but it is correct and the approach that I use myself.

Inconsistent state information

Creating a Blu-ray image is challenging to Encore. More often than not, the BD creation will fail with error messages such as:

• Blu-ray Error: “device not ready”, Code: “3″, Note “File can not open.(…\AuthorScriptHDMVSessions\….ac3.ves)
• Blu-ray Object: “nameoftimeline”, Error: “invalid parameter”, Code: “19″, Note: “Failed in Open of the VES file”
• Blu-ray Object: “nameofmenu”, Export file not found

To work around these errors, either delete the .ves file in question, or take out the big guns and delete the AuthorScriptHDMVSessions (and the Cache subdirectories) from the project directory. This causes Encore to re-analyze all of the assets and clear up the confusion.

Bugs, bugs and nothing but bugs

Authoring dual layer discs may cause problem such as:

• Blu-ray Error: “Not enough free storage”, Code: “25″, Note: “Not enough space on media” on 50G DL BD-R.
• Blu-ray Error: “file already exists”, Code: “6″, Note “10485888 10489952 11826176″

Cache\HDMVProjectExport.log reveals “Not enough space on media, layer 0 size greater than max layer size”. Encore estimated the size of .m2ts files wrong, and as a result it can not fit the content on the first layer. See forums.adobe.com. A closer look at the .log will point you to the file that it has problems placing. Some times adding/deleting some BD-ROM Content, or re-encoding at a different bitrate appears to work around the problem. Alternately one can use “burn to folder” and then create an .iso using imgburn

Subtitle woes

Error messages that fall into this category are

• Timecode overlap found in subtitle script line
• Blu-ray error encore cs5 code 13 errors in some PGs (when you told Encore to ignore the gab requirement)
• Errors in Some PGs: end time of preceding display set is too close to the start time of succeeding display set, see trace log for more information

There should be a gap of about 3 to 5 frames between subtitles. This appears to be a Encore specific “feature”. To work around this, use the scripting engine from Subtitle Workshop 4 to create these gap. Remember to set the correct frame rate in Subtitle Workshop! Note that in some rare circumstances this script truncates subtitles to 0 length when there is no space for them.

(* --------------------------------------------------------------------------
 *               Enforce 5 frame gap for Adobe Encore subtitles
 *                    Subtitle Workshop 4 - Pascal scripts
 *                          http://www.coertvonk.com
 *
 * Adobe Encore is very picky about enforcing a gap between subtitles.
 * This routine will enforce a 222 msec gap.  For 24p that corresponds to
 * about 5 frames.
 *
 * This script prevents errors like "subtitle too close" in Encore
 *
 * Remember to set the correct frame rate (FPS) before calling this script
 *
 * Usage: place this script in the "Subtitle Workshop\PascalScripts"
 *        directory.  Run the script from Subtitle Workshop via
 *        Tools/Pascal scripts menu, Tools > Pascal scripts > BlurayGap.pas
 *)

program BlurayGap;

var
   ii	     : Integer;
   Count     : Integer;
   start     : Integer;
   final     : Integer;
   startNext : Integer;
   gap       : Integer;

begin
   gap := 222; // [msec]
   Count := GetSubtitleCount;<br />

   for ii := 0 to Count-2 do
   begin
      start := GetSubtitleInitialTime(ii);
      final := GetSubtitleFinalTime(ii);
      startNext := GetSubtitleInitialTime(ii+1);

      if final + gap > startNext then
      begin
	 if startNext > start + gap then
	    SetSubtitleFinalTime(ii, startNext - gap)
	 else
	    SetSubtitleFinalTime(ii, start); (* should really delete the subtitle .. *)
      end;
   end;

end.

• voip.lan = SPA2102 VoIP Adapter connected to the LAN
• rtr.lan = DD-WRT firmware router, LAN interface
• rtr.wan = DD-WRT firmware router, WAN interface
• sipsorcery.com
• 4153767253@podlinez.net = CNN headlines playback

Registration

1. voip.lan -> rtr.lan

   REGISTER sip: sipsorcery.com SIP/2.0
   Via: SIP/2.0/UDP voip.lan:5060;branch=z9hG4bK-1b604b85;rport
   Contact: User Name <sip:username@voip.lan:5060>;expires=3600

2. rtr.wan -> sipsorcery.com
   

   REGISTER sip:sipsorcery.com SIP/2.0
   Via: SIP/2.0/UDP rtr.wan;branch=z9hG4bKe517.cd4209f5.0
   Via: SIP/2.0/UDP voip.lan:5060;branch=z9hG4bK-1b604b85;rport=5060
   Contact: User Name <sip:username@rtr.wan:5060>;expires=3600

3. sipsorcery.com -> rtr.wan
   

   SIP/2.0 401 Unauthorised
   Via: SIP/2.0/UDP rtr.wan;branch=z9hG4bKe517.cd4209f5.0;received=rtr.wan;rport=5060
   Via: SIP/2.0/UDP voip.lan:5060;branch=z9hG4bK-1b604b85;rport=5060
   WWW-Authenticate: Digest realm=”sipsorcery.com”,nonce=”11086481671708617043″

4. rtr.lan -> voip.lan
   

   SIP/2.0 401 Unauthorised
   Via: SIP/2.0/UDP voip.lan:5060;branch=z9hG4bK-1b604b85;rport=5060
   WWW-Authenticate: Digest realm=”sipsorcery.com”,nonce=”11086481671708617043″

5. voip.lan -> rtr.lan (this time with MD5 authentication)
   

   REGISTER sip: sipsorcery.com SIP/2.0
   Via: SIP/2.0/UDP voip.lan:5060;branch=z9hG4bK-5d9fc9b1;rport
   Authorization: Digest username=”username”,realm=”sipsorcery.com”,nonce=”11086481671708617043″,uri=”sip:sipsorcery.com”,algorithm=MD5,response=”8d6081d09dc9ce406e7d541317ef9f6d”
   Contact: User Name <sip:username@voip.lan:5060>;expires=3600

6. rtr.wan -> sipsorcery.com (send twice)
   

   REGISTER sip:sipsorcery.com SIP/2.0
   Via: SIP/2.0/UDP rtr.wan;branch=z9hG4bKb517.d7fa54b4.0
   Via: SIP/2.0/UDP voip.lan:5060;branch=z9hG4bK-5d9fc9b1;rport=5060
   Authorization: Digest username=”username”,realm=”sipsorcery.com”,nonce=”11086481671708617043″,uri=”sip:sipsorcery.com”,algorithm=MD5,response=”8d6081d09dc9ce406e7d541317ef9f6d”
   Contact: User Name <sip:username@rtr.wan:5060>;expires=3600

7. sipsorcery.com -> rtr.wan (send twice)
   

   SIP/2.0 200 Ok
   Via: SIP/2.0/UDP rtr.wan;branch=z9hG4bKb517.d7fa54b4.0;received=rtr.wan;rport=5060
   Via: SIP/2.0/UDP voip.lan:5060;branch=z9hG4bK-5d9fc9b1;rport=5060
   Contact: <sip:username@rtr.wan:5060;rinstance=dd126b5df19eaccd>;expires=3424,<sip:username@rtr.wan:5060>;expires=3600

8. rtr.lan -> voip.lan
   

   SIP/2.0 200 Ok
   Via: SIP/2.0/UDP voip.lan:5060;branch=z9hG4bK-5d9fc9b1;rport=5060
   Contact: <sip:username@rtr.wan:5060;rinstance=dd126b5df19eaccd>;expires=3424,<sip:username@rtr.wan:5060>;expires=3600

No-op

1. sipsorcery.com -> rtr.wan (every 10 seconds)

   UDP Src Port: sip (5060), Dst Port: sip (5060)
   Data (4 bytes):
   00 00 00 00

Phone dials “266″ and the VoIP adapter generates an INVITE

1. voip.lan -> rtr.lan:5060

   INVITE sip: 266@sipsorcery.com SIP/2.0
   Via: SIP/2.0/UDP voip.lan:5060;branch=z9hG4bK-56847992;rport
   Remote-Party-ID: User Name <sip:username@sipsorcery.com>;screen=yes;party=calling
   Contact: User Name <sip:username@voip.lan:5060>
   SDP:
   Owner/Creator, Session Id (o): – 3936 3936 IN IP4 voip.lan
   Connection Information (c): IN IP4 voip.lan
   m=audio 16462 RTP/AVP 18 8 0 2 4 96 97 98 100 101

rtr starts RTP proxy configuration, replies with TRYING and forwards the INVITE

1. rtr

   proxy RTP traffic from
   (src=voip.lan:16462, dst=rtr.lan:38826) -> (src=rtr.wan:35740, dst=???), and visa versa
   details:
   new session 7e641be2-5ef40765@voip.lan, tag c4fbd9be47ee0339o0;1 requested, type strong
   new session on a port 35740 created, tag c4fbd9be47ee0339o0;1
   pre-filling caller’s address with voip.lan:16462

2. rtr.lan -> voip.lan
   

   SIP/2.0 100 trying – your call is important to us
   Via: SIP/2.0/UDP voip.lan:5060;branch=z9hG4bK-56847992;rport=5060
   Warning: 392 rtr.lan:5060 “Noisy feedback tells: pid=15874 req_src_ip=voip.lan req_src_port=5060 in_uri=sip:266@sipsorcery.com out_uri=sip:4153767253@podlinez.net via_cnt==1″

3. rtr.wan -> podlinez.net (with RTP transport address changed from voip.lan:16462 to rtr.wan:35740)
   

   INVITE sip:4153767253@podlinez.net SIP/2.0
   Record-Route: <sip:rtr.wan;r2=on;ftag=c4fbd9be47ee0339o0;lr=on>
   Record-Route: <sip:rtr.lan;r2=on;ftag=c4fbd9be47ee0339o0;lr=on>
   Via: SIP/2.0/UDP rtr.wan;branch=z9hG4bKf36d.99615505.0
   Via: SIP/2.0/UDP voip.lan:5060;branch=z9hG4bK-56847992;rport=5060
   Remote-Party-ID: User Name <sip:username@sipsorcery.com>;screen=yes;party=calling
   Contact: User Name <sip:username@rtr.wan:5060>
   SDP:
   Owner/Creator, Session Id (o): – 3936 3936 IN IP4 voip.lan
   Connection Information (c): IN IP4 rtr.wan
   Media Description, name and address (m): audio 35740 RTP/AVP 18 8 0 2 4 96 97 98 100 101

podlinez.net responds with TRYING and OK

1. podlinez.net -> rtr.wan

   SIP/2.0 100 Trying
   Via: SIP/2.0/UDP rtr.wan;branch=z9hG4bKf36d.99615505.0;received=rtr.wan
   Via: SIP/2.0/UDP voip.lan:5060;branch=z9hG4bK-56847992;rport=5060
   Record-Route: <sip:rtr.wan;r2=on;ftag=c4fbd9be47ee0339o0;lr=on>
   Record-Route: <sip:rtr.lan;r2=on;ftag=c4fbd9be47ee0339o0;lr=on>
   Contact: <sip:4153767253@podlinez.net>

2. podlinez.net -> rtr.wan
   

   SIP/2.0 200 OK
   Via: SIP/2.0/UDP rtr.wan;branch=z9hG4bKf36d.99615505.0;received=rtr.wan
   Via: SIP/2.0/UDP voip.lan:5060;branch=z9hG4bK-56847992;rport=5060
   Record-Route: <sip:rtr.wan;r2=on;ftag=c4fbd9be47ee0339o0;lr=on>
   Record-Route: <sip:rtr.lan;r2=on;ftag=c4fbd9be47ee0339o0;lr=on>
   Contact: <sip:4153767253@podlinez.net>
   SDP:
   Session Description Protocol Version (v): 0
   Owner/Creator, Session Id (o): root 57052 57052 IN IP4 podlinez.net
   Connection Information (c): IN IP4 podlinez.net
   Media Description, name and address (m): audio 14684 RTP/AVP 0 101

rtr completes proxy configuration and forwards OK

1. rtr

   proxy RTP traffic from
   (src=voip.lan:16462, dst=rtr.lan:38826) -> (src=rtr.wan:35740, dst=podlinez.net:14684), and visa versa
   details:
   lookup on ports 35740/38826, session timer restarted
   pre-filling callee’s address with 208.68.167.230:14684

2. rtr.lan -> voip.lan (with the RTP transport address changed from podlinez.net:14684 to rtr.lan:38826)
   

   SIP/2.0 200 OK
   Via: SIP/2.0/UDP voip.lan:5060;branch=z9hG4bK-56847992;rport=5060
   Record-Route: <sip:rtr.wan;r2=on;ftag=c4fbd9be47ee0339o0;lr=on>
   Record-Route: <sip:rtr.lan;r2=on;ftag=c4fbd9be47ee0339o0;lr=on>
   Contact: <sip:4153767253@podlinez.net>
   SDP:
   Owner/Creator, Session Id (o): root 57052 57052 IN IP4 podlinez.net
   Connection Information (c): IN IP4 rtr.lan
   Media Description, name and address (m): audio 38826 RTP/AVP 0 101

voip.lan sends ACK

1. voip.lan -> rtr.lan

   ACK sip: 4153767253@podlinez.net SIP/2.0
   Via: SIP/2.0/UDP voip.lan:5060;branch=z9hG4bK-417fc59a;rport
   Route: <sip:rtr.lan;r2=on;ftag=c4fbd9be47ee0339o0;lr=on>, <sip:rtr.wan;r2=on;ftag=c4fbd9be47ee0339o0;lr=on>
   Contact: User Name <sip:username@voip.lan:5060>

2. rtr.wan -> podlinez.net
   

   ACK sip:4153767253@podlinez.net SIP/2.0
   Record-Route: <sip:rtr.wan;r2=on;ftag=c4fbd9be47ee0339o0;lr=on>
   Record-Route: <sip:rtr.lan;r2=on;ftag=c4fbd9be47ee0339o0;lr=on>
   Via: SIP/2.0/UDP rtr.wan;branch=0
   Via: SIP/2.0/UDP voip.lan:5060;branch=z9hG4bK-417fc59a;rport=5060
   Contact: User Name <sip:username@voip.lan:5060>
   P-hint: rr-enforced

Audio (RTP) streams establish

1. rtr.wan:35740 -> podlinez.net:14684 (and so on ..)

   RTP
   PT=ITU-T G.711 PCMU, SSRC=0x8E84822D, Seq=7574, Time=460296833

2. podlinez.net:14684 -> rtr.wan:35740 (starts about .6 sec later)
   

   RTP
   PT=ITU-T G.711 PCMU, SSRC=0x2C0B6B9C, Seq=60649, Time=160

Phone hangs up, and VoIP adapter sends BYE

1. voip.lan -> rtr.lan

   BYE sip: 4153767253@podlinez.net SIP/2.0
   Via: SIP/2.0/UDP voip.lan:5060;branch=z9hG4bK-6939ac8f;rport
   Route: <sip:rtr.lan;r2=on;ftag=c4fbd9be47ee0339o0;lr=on>, <sip:rtr.wan;r2=on;ftag=c4fbd9be47ee0339o0;lr=on>

2. rtr.wan -> podlinez.net
   

   BYE sip:4153767253@podlinez.net SIP/2.0
   Record-Route: <sip:rtr.wan;r2=on;ftag=c4fbd9be47ee0339o0;lr=on>
   Record-Route: <sip:rtr.lan;r2=on;ftag=c4fbd9be47ee0339o0;lr=on>
   Via: SIP/2.0/UDP rtr.wan;branch=z9hG4bKc36d.1d7216.0
   Via: SIP/2.0/UDP voip.lan:5060;branch=z9hG4bK-6939ac8f;rport=5060
   P-hint: rr-enforced

3. podlinez.net -> rtr.wan
   

   SIP/2.0 200 OK
   Via: SIP/2.0/UDP rtr.wan;branch=z9hG4bKc36d.1d7216.0;received=rtr.wan
   Via: SIP/2.0/UDP voip.lan:5060;branch=z9hG4bK-6939ac8f;rport=5060
   Record-Route: <sip:rtr.wan;r2=on;ftag=c4fbd9be47ee0339o0;lr=on>
   Record-Route: <sip:rtr.lan;r2=on;ftag=c4fbd9be47ee0339o0;lr=on>

4. rtr.lan -> voip.lan
   

   SIP/2.0 200 OK
   Via: SIP/2.0/UDP voip.lan:5060;branch=z9hG4bK-6939ac8f;rport=5060
   Record-Route: <sip:rtr.wan;r2=on;ftag=c4fbd9be47ee0339o0;lr=on>
   Record-Route: <sip:rtr.lan;r2=on;ftag=c4fbd9be47ee0339o0;lr=on>

• Session Initiation Protocol  (RFC3261) and (RFC3581), for call control
• Session Description Protocol (RFC 4566 and RFC4961) for audio stream establishment
• Real Time Protocol (RTP) for the audio streams (basically adds a time-stamp to UDP)

The client will include its address and port in three different places (src)

• In the Via header of a request, as the target of the response
• In the Contact of a SIP REGISTER message, as the target for an incoming SIP INVITE
• In the SDP, as the target of the audio connection

Public IP Address

Phone registering

1. When a SIP phone client sends a REGISTER request to a DID server, this message contains a Via header field with a send-by parameter that contains the address/port on the client.  As we will see the header allow the response to traverse the same path.  The Contact field informs the server how to reach the  SIP phone for i.e. incoming calls.  An example:
   

   Via: SIP/2.0/UDP 67.170.142.180:5060
   Contact: User Name <sip:username@67.170.142.180:5060>

2. The server (and each proxy) extends the top-most Via header with the address that they received the message from (“Reflexive Transport Address”), and then add their own Via header to the top of the message.  This is how the server learns about the message path.  From the contact field it learns how to contact the SIP phone directly.  I.e. for an incoming call (INVITE request).  An example:
   

   Via: SIP/2.0/UDP sip.sipsorcery.com:5060
   Via: SIP/2.0/UDP proxy.somedomain.com:5060;received=proxy.somedomain.com
   Via: SIP/2.0/UDP 67.170.142.180:5060;received=67.170.142.180
   Contact: User Name <sip:username@67.170.142.180:5060>

3. When the server replies, it starts with the same set of Via headers.  The message will be sent from the addr/port listed in the top-most Via parameter.  It then removes this top-most Via and sent the message to addr/port listed in the then top-most Via parameter.  When the client finally receives the response it will have its own Via header, with the received added.  For example:
   

   Via: SIP/2.0/UDP 67.170.142.180:5060;received=67.170.142.18

Phone placing a call

1. When the SIP phone client sends an INVITE request to the DID server, the Via headers will be used as with the REGISTER request.  The message body will also include the session description with the addr/port where it listens for the audio connection:
   

   Via: SIP/2.0/UDP 67.170.142.180:5060
   Contact: User Name <sip:username@67.170.142.180:5060>
   Message Body (SDP):

   Connection Information (c): IN IP4 67.170.142.180
   Media Description, name and address (m): audio 16422 RTP/AVP 18 8 0 2 4 96 97 98 100 101

2. The server will first reply with response 180 (Trying), and the SIP phone will sound the ringing tone.  The server then replies with response 200 (OK) and include the session description with the addr/port where it listens for the audio connection.  The client receives:
   

   Via: SIP/2.0/UDP 67.170.142.180:5060;received=67.170.142.180
   Contact: <sip:174.129.236.7:5060>
   Message Body (SDP)

   Connection Information (c): IN IP4 208.68.167.230
   Media Description, name and address (m): audio 17904 RTP/AVP 0 101

3. The SIP phone sends an ACK, and builds the RTP audio path from 67.170.142.180:16422 to 208.68.167.230:17904.

Phone receiving a call

When the DID receives an call over the PSTN, it generates a INVITE message to the SIP phone.  The URI that it uses for the phone is what it learned from the Contact field in the REGISTER message.  The call proceeds similar to the outgoing call described above.

Network Address Translation

Phone registering

1. The client populates the top-most Via header field.  As usual the send-by parameter contains its IP address and port number.  It also adds an “rport” parameter with no value to request the receive-port value from the next hop.  An example:
   

   Via: SIP/2.0/UDP 10.0.1.10:5060;rport
   Contact: User Name <sip:username@10.0.1.10:5060>

2. The NAT forwards the packet from an external port on its public IP address.
3. The server (and each proxy) fill in the “received” and “rport” values in the Via header to indicate the source of the message.  It then add its own Via header to the top of the message.  An example:
   

   Via: SIP/2.0/UDP sip.sipsorcery.com:5060
   Via: SIP/2.0/UDP proxy.somedomain.com:5060;rport=5060;received=proxy.somedomain.com
   Via: SIP/2.0/UDP 10.0.1.10:5060;rport=1024;received=67.170.142.180
   Contact: User Name <sip:username@10.0.1.10:5060>

4. When the server replies, it starts with the same set of Via headers.  The message will be sent from the addr/port listed in the top-most Via parameter.  It then removes this top-most Via parameter and sends the message to the addr/port listed in the then top-most Via parameter.  When the client finally receives the response, the only Via header that is left is the one that it inserted itself, but with the received/rport reflecting public IP addr/port on the NAT router.  For example:
   

   Via: SIP/2.0/UDP 10.0.1.10:5060;rport=1024;received=67.170.142.180

5. The client now knows its public IP addr/port and re-registers itself with this addr/port in the Contact field.  This way the DID knows how to contact the SIP phone when it receives an incoming call.  Again, this relies on the NAT binding being kept active.
   

   Via: SIP/2.0/UDP 67.170.142.180:1024;rport
   Contact: User Name <sip:username@67.170.142.180:1024>

6. In our case sipsorcery.com is configured to keep the NAT binding active by sending no-op messages to the external port 67.170.142.180:1024.

Phone placing a call

1. When the SIP phone client sends an INVITE request to the DID server, the Via headers will be similar as what we saw in the REGISTER request.  The message body will also include the session description with its addr/port where it listens for the audio connection:
   

   Via: SIP/2.0/UDP 67.170.142.180:5060
   Contact: User Name <sip:username@67.170.142.180:9988>
   Message Body (SDP):

   Connection Information (c): IN IP4 67.170.142.180
   Media Description, name and address (m): audio 16422 RTP/AVP 18 8 0 2 4 96 97 98 100 101

2. The server will first reply with response 180 (Trying), and the SIP phone will sound the ringing tone.  It then replies with response 200 (OK) and include the SDP.  The client receives:
   

   Via: SIP/2.0/UDP 67.170.142.180:5060;received=67.170.142.180
   Contact: <sip:174.129.236.7:5060>
   Message Body (SDP)

   Connection Information (c): IN IP4 208.68.167.230
   Media Description, name and address (m): audio 17904 RTP/AVP 0 101

3. The SIP phone sends an ACK, builds the RTP audio path from 67.170.142.180:16422 to 208.68.167.230:17904.  Assuming the server uses symmetric RTP, the reverse audio path will traverse the NAT and reach the SIP phone.

As you see, the SDP Media Description lists an internal port number that is not reachable from the outside.  One work-around is to create a static port range forwarding in the NAT.  A more elegant solution is that modern SIP/SDP implementations ignore the RTP port in the SDP payload and wait until they receive media packets from the NAT’d client.  It then look for the actual source port of the incoming RTP stream and start sending media there, instead.  Another solution is to use Application Level Gateways, but they rarely work correctly.  Yet another solution based on OpenSER (MilkFish) will be described later.

Phone receiving a call

When the DID receives an call over the PSTN, it generates a INVITE message to the SIP phone.  The URI that it uses for the phone is what it learned from the Contact field in the REGISTER message.  The call proceeds similar to the outgoing call described above.  Note that the audio from the DID to the SIP Phone will only traverse the NAT after the SIP Phone builds the audio path to the DID.

Outbound proxy

Here we will use MilkFish (OpenSER) as the outbound proxy.  Instead of having to go through the NAT, the phones now go through the proxy service on the router.  Note that the rport option is not required.

Phone registering

1. The client populates the top-most Via header field.  As usual the send-by parameter contains its IP address and port number.  It also adds an “rport” parameter with no value to request the receive-port value from the next hop.  An example:
   

   Via: SIP/2.0/UDP 10.0.1.10:5060;rport
   Contact: User Name <sip:username@10.0.1.10:5060>

2. The outbound proxy adds its own “Via” line and corrects the “Contact:”.  It then forwards the message on the router’s public IP address.
   

   Via: SIP/2.0/UDP 67.170.142.180
   Via: SIP/2.0/UDP 10.0.1.10:5060;rport=5060
   Contact: User Name <sip:username@67.170.142.180:5060>

3. The server add its own Via header to the top of the message.
   

   Via: SIP/2.0/UDP sip.sipsorcery.com:5060
   Via: SIP/2.0/UDP 67.170.142.180
   Via: SIP/2.0/UDP 10.0.1.10:5060;rport=5060
   Contact: User Name <sip:username@67.170.142.180:5060>

4. When the server replies, it copies the headers from the request.  The response will be sent from the transport address listed in the top-most Via parameter.  It then removes this top-most Via parameter and sends the message to the transport address listed in the next Via parameter.  The proxy server does the same, and when the client finally receives the response, the only Via header that is left is the one that it inserted itself, except that the domain name reflects the public IP address of the outbound proxy.
   

   Via: SIP/2.0/UDP 10.0.1.10:5060;rport=5060
   Contact: User Name <sip:username@67.170.142.180:5060>

Phone placing a call

1. When the SIP phone client sends an INVITE request to the DID server, the Via headers will be similar as what we saw in the REGISTER request.  The message body will also include the session description with the transport address where it listens for the audio connection:
   

   Via: SIP/2.0/UDP 67.170.142.180:5060
   Contact: User Name <sip:username@67.170.142.180:9988>
   Message Body (SDP):Connection Information (c): IN IP4 67.170.142.180
   Media Description, name and address (m): audio 16422 RTP/AVP 18 8 0 2 4 96 97 98 100 101

2. @@
   

   Via: SIP/2.0/UDP 67.170.142.180:5060
   Contact: User Name <sip:username@67.170.142.180:9988>
   Message Body (SDP):Connection Information (c): IN IP4 67.170.142.180
   Media Description, name and address (m): audio 16422 RTP/AVP 18 8 0 2 4 96 97 98 100 101

3. @@
   

   Via: SIP/2.0/UDP 67.170.142.180:5060
   Contact: User Name <sip:username@67.170.142.180:9988>
   Message Body (SDP):Connection Information (c): IN IP4 67.170.142.180
   Media Description, name and address (m): audio 16422 RTP/AVP 18 8 0 2 4 96 97 98 100 101

4. @@
   

   Via: SIP/2.0/UDP 67.170.142.180:5060
   Contact: User Name <sip:username@67.170.142.180:9988>
   Message Body (SDP):Connection Information (c): IN IP4 67.170.142.180
   Media Description, name and address (m): audio 16422 RTP/AVP 18 8 0 2 4 96 97 98 100 101

5. @@
   

   Via: SIP/2.0/UDP 67.170.142.180:5060
   Contact: User Name <sip:username@67.170.142.180:9988>
   Message Body (SDP):Connection Information (c): IN IP4 67.170.142.180
   Media Description, name and address (m): audio 16422 RTP/AVP 18 8 0 2 4 96 97 98 100 101

6. @@
   

   Via: SIP/2.0/UDP 67.170.142.180:5060
   Contact: User Name <sip:username@67.170.142.180:9988>
   Message Body (SDP):Connection Information (c): IN IP4 67.170.142.180
   Media Description, name and address (m): audio 16422 RTP/AVP 18 8 0 2 4 96 97 98 100 101

7. @@
   

   Via: SIP/2.0/UDP 67.170.142.180:5060
   Contact: User Name <sip:username@67.170.142.180:9988>
   Message Body (SDP):Connection Information (c): IN IP4 67.170.142.180
   Media Description, name and address (m): audio 16422 RTP/AVP 18 8 0 2 4 96 97 98 100 101

Introduction

Google Voice (GV) gives us free domestic calling.  It also forks incoming calls to one or more phones.  It is implemented as an interconnect for the traditional phone system (PSTN).  It routes the calls as follows:

• Incoming calls: are forked to one or more phones.  I.e. your land line phone and a cell phone.  If the call is left unanswered, it rolls over to voice mail that it transcribes.
• Outgoing calls: you setup a call through their web interface or by dialing your own number.  GV will first call you, and then call the party that you are calling.  It then cross connect the two.

While this approach works well, people like me want to not only free themselves from the long distance companies, but also from their land line phone.  However, at the time of this writing, Google Voice does not directly support VoIP.

We will work around this by using a VoIP/PSTN interconnect such as netgate.com.  This is known as Direct inward Dialing (DID).  The DID gives you a regular phone number that you can use with your VoIP phone.  Examples of VoIP phones are hardware SIP phones, software SIP phones such as x-lite, or regular phones hooked up through an Analog Telephone Adapters (ATA).  In this article we will look at the SPA2102 analog phone adapter.  Here we register our DID phone number as one of the phones with Google Voice.

Call routing using GV + SipGate is as follows:

• Incoming calls: arrive at Google voice.  They will forward the call to your DID phone number that then rings your VoIP connected phone.
• Outgoing calls: you setup a call through the Google Voice  web interface or by dialing your own number.  GV first calls your DID number that in turn rings your VoIP phone.  GV then calls the party that you are calling and connect the two calls.

The next step make our phones behave like regular phones when dialing out.  Instead of going through the web interface, we want to dial a number on the phone.  This can be done by enlisting yet-another service: sipsorcery.com.  This is called a called a back-to-back user agent (B2BUA).  This service sits in between your analog phone adapter (SPA2102) and the DID (sipgate.com).

A typical call is visualized in this call graph.  Call routing using GV + SipGate + SipSorgery is as follows:

• Incoming calls: arrive at Google voice.  They will forward the call to your DID phone number (sipgate.com) that in turn forwards the call to sipsorcery.com that then rings your VoIP connected phone.
• Outgoing calls: you just dial the number on your VoIP connected phone.  Sipsourcery requests Google voice to setup the call.

Drawbacks

• With so many services involved, there are many points of failure.
• Another drawback is that 911 may not work or at best will not automatically get your location.  You will need to adjust the dialing script on sipsorcery to route the calls through netgate.com and enable/pay for E911 service.

Edge router: DD-WRT firmware

Start by verifying that your network connection is VoIP-ready by running this test.

First things first, we need connectivity.  I use a Asus RT-N16 with DD-WRT firmware configured as follows:

• Setup > Basic Setup
  • WAN Setup > Optional Settings
    • Router Name =rtr
    • Domain Name = vonk
  • Network Setup > Router IP
    • Local IP Address = 10.0.1.1
    • Subnet Mask = 255.255.255.0
  • Network Setup > Network Address Server Settings (DHCP)
    • DHCP Type = DHCP Server
    • DHCP Server = Enable
    • Start IP Address = 10.0.1.100
    • Static DNS 1 =8.8.8.8  # google public DNS
    • Static DNS 2 = 8.8.4.4
  • Network Setup > Time Settings (man)
    • Time Zone = UTC-8
    • Summer Time (DST) = 2nd Sun Mar – first Sun Nov
• Setup > DDNS
  • DDNS
    • DDNS Service = no-ip.com
    • User Name = username
    • Password = password
    • Hostname = hostname
    • Do not use external ip check = no
• Services > Services
  • DHCP Server (man)
    • LAN Domain = your sirname
    • Static Leases
      • press “Add”, enter the information, press “Save”, repeat until all added
      • press “Apply Settings”
  • DNSMasq (man)
    • DNSMasq = enable
    • Local DNS = enable
    • Additional DNSMasq Options
      • # Additional options
        domain-needed                    # don’t forward plain names
        bogus-priv                       # don’t forward private addresse
        bogus-nxdomain=64.94.110.11      # keep Verisign in control
        filterwin2k                      # filter useless Windows DNS requests
        # Local DNS name server
        local=/yoursirname/    # answer this domain from /etc/hosts
        expand-hosts    # add the domain to /etc/hosts entries
        dhcp-option=2,-28800             # UTC -8:00
        dhcp-option=7,10.0.1.202         # SYSLOG server
        #dhcp-option=42,10.0.1.1          # NTP server
        dhcp-option=43,01:04:00:00:00:02 # disable NetBIOS over TCP/IP
  • Secure Shell (man)
    • Password Login = false
    • Authorized Keys = paste your public ssh-rsa key

• Wireless
  • Basic Settings
    • Wireless Physical Interface wl0
      • Wireless Mode = AP
      • Wireless Network Mode = NG-Mixed
      • Wireless Network Name = your SSID
      • Wireless Channel = auto
      • Channel Width = 40 MHz
      • Control Channel = upper
      • Network Configuration = bridged
    • Virtual Interfaces wl0.1
      • Wireless Network Name = public
      • AP Isolation = enable
      • Network Configuration = bridged
  • Wireless Security
    • Physical Interface wl0
      • Security Mode = WPA2 Personal
      • WPA Algorithms = AES
      • WPA Shared Key = your wpa shared key
    • Virtual Interfaces wl0.1
      • Security Mode = disable
• Setup > Networking (man)
  • Bridging
    • Create Bridge
      • press “Add” to add bridge “br1″
      • press “Save”
      • IP Address  =10.0.0.1
      • Subnet Mask = 255.255.255.0
      • press “Apply Settings”, and wait a few seconds for “br1″ to show up in “Current Bridging Table”
    • Assign to Bridge
      • press “Add” to add bridge “br1″ on interface wl0.1
      • press “Apply Settings”
  • DHCPD
    • Multiple DHCP Server
      • br1, on, 100, 50, 3600
      • press “Apply Settings”
• Administration > Management
  • Web Access
    • Enable Info Site = enable
    • Info Site MAC Masking = enable
• Access Restrictions > WAN Access
  • Blocked Services
    • smtp
• NAT/QoS > Port Forwarding
  • Forwards
    • press “Add”, enter the information, press “Save”, repeat until all added
    • press “Apply Settings”

Sipgate.com (DID) and Sipsorcery.com (B2BUA)

Details can be found in section 8 of Unlimited free calling with Google Voice.   Remember that your sipgate VoIP credentials are not your username and password.  Instead use sipgate.com > My Settings > SIP Credentials.

VoIP adapter (SPA2102)

Some popular analog telephone adapters (ATA) from Cisco/Linksys are

• PAP2T-NA, an old time favorite that provides the basic functionality: VoIP adapter with two analog phone hookups.
• SPA2102, same as PAP2T-NA, but adds a router
• SPA3102, same as SPA2102, but supports a land line for fallback connection, also support multiple SIP accounts per line.

My configuration:

• Voice > Line 1
  • SIP Settings
    • SIP Transport = UDP (default)
    • SIP Port = 5060 (default)
  • Network Settings
    • SIP ToS/DiffServ Value = 0×04 (Type-of-Service = high reliability)
    • RTP Tos/DiffServ Value = 0×10 (Type-of-Service = minimal delay)
    • Network Jitter Level = low (assuming that is true for your connection)
  • Proxy and Registration
    • Proxy = sip.sipsorcery.com
    • Register = yes
  • Audio Configuration
    • Silence Supp Enable = false (otherwise the other party notices the cutting in/out)
  • Proxy and Registration
    • Use DNS SRV = yes (use DNS to switch between SIP servers) (details) (def)
    • DNS SRV Auto Prefix = auto

• Voice > SIP
  • RTP Parameters
    • RTP Packet Size = 0.020 sec (minimize delay at the cost of some overhead)

Examples of event sequences where the SIP phone has a route-able IP address can be found in Transport address in SIP and SDP messages.

VoIP adapter (SPA2102) behind a NAT router

The most logical place for the VoIP adapter would be in between the cable modem and the edge router.  This way it can prioritize VoIP traffic and keep security issues limited to the VoIP adapter itself.  However, the SPA2102 has some flaws.  The most obvious one affects active SCP and TFTP transfers.  It also maxes out at about 20 Mbps with small packet throughput far worse.

The alternative is to place the VoIP adapter behind the Edge Router.  This router very likely implements Network Address Translation (NAT) to share one public IP address with many local devices.  However, the SIP protocol predates NAT and requires some tweaks to make it work.  One of the tweaks is formalized in RFC3581 and extends SIP with support for outgoing calls over NAT.  In this the client adds an “rport” parameter to the Via header field.

The first configuration step should be to creates a static DHCP lease for the VoIP adapter, and enable web admin access.

• On the DD-WRT router
  • Services > Services > Services Management > DHCP Server
    • add a static lease for the VoIP adapter’s WAN address, so we can easily access it

• On the VoIP adapter
  • Voice > System
    • System Configuration
      • Web Admin Access = yes (also, remember to set the passwords)

As described in Transport address in SIP and SDP messages, placing the  VoIP adapter behind a NAT can be made to work, but requires the DID to support Symmetric SIP (RFC3581) and Symmetric RTP (RFC4961).

Signaling (SIP) behind NAT

There are several solutions to make the signaling protocol aware of network address translation:

solution 1

The signaling protocol (SIP, RFC5626 ) has a build-in mechanism to help traverse NAT.   This solution takes advantage of this mechanism.  In that the VoIP adapter learns the address and port on the external NAT interface through Via parameters.

• On the VoIP adapter
  • Voice > Line 1
    • NAT Settings
      • NAT Mapping Enable = yes (use external IP address as the SIP Contact)
      • NAT Keep Alive Enable = no (my DID already does that for me)
  • Voice > SIP
    • NAT Support Parameters (documentation)
      • Handle VIA received = yes  (learn the public IP address from a received “Via: ….; received=publicIpAdd“)
      • Handle VIA rport = yes  (learn the public port number from a received “Via: ….; rport=publicUdpPort“)
      • Substitute VIA Addr = yes  (add “Via: …; received=remoteIpAddr)
      • Send Resp to Src Port = yes  (add “Via: …; rport=remoteUdpPort)
      • EXT RTP Port Min = (left blank)
      • NAT Keep Alive Intvl = 15 sec  (maximum inactive time for the NAT binding)

solution 2: SIP port forwarding

In addition to above described configuration, you can help SIP pass through the NAT by adding a static forwarding of port range 5060-5061.  This works even when the DID does not support the “VIA rport”.  It also allows for direct client to client calling because the external port number is static.  This however requires a similar port range to be forwarded for each SIP phone.

• On the edge router
  • NAT/QoS > Port Range Forward
    • voip sip, 5060-5061, TCP and UDP, 10.0.1.10 (or whatever the addr for the VoIP adapter)

• On the VoIP adapter
  • Voice > Line 1
    • SIP Settings
      • SIP Ext Port = 5060

solution 3: outbound proxy

Using a outbound proxy helps SIP bypasses the NAT and requires minimal configuration on the VoIP adapter.  The DD-WRT router contains the OpenSER  “MilkFish” service that works well.  It also allows for local registration of SIP phones.  For an example message trace refer to SIP/SDP trace using MilkFish as outbound proxy.  Remember to disable STUN on all the phones in the local network.  This to prevent one of these phones from grabbing external port 5060.  Alternatively, you can use the firewall to enforce this behavior (see this blog)

solution 4: ALG

Another approach that comes to mind is Application Level Gateways (ALG), but I have not seen one that functions correctly.

Media streams (SDP/RTP) behind NAT

Now that we have the signaling path taken care of, it is time to turn our attention to the media streams.  SIP messages that setup the RTP media path use the media description protocol (SDP, RFC 4566).  This SDP is not NAT-aware, and can only be used behind NAT with the help of other protocols.

solution 1

This solution assumes that the DID is “smart” about NAT.  By “smart”, I mean that it ignores the RTP port in the SDP payload and waits until it receives media packets from the NAT’d client.  It then look for the actual source port of the incoming RTP stream and start sending media there, instead.

solution 2: RTP port forwarding

If the DID is not “smart” about NAT, you add a static forwarding of external port range 16384-16482 for RTP.  This however requires the configuration on the NAT router for each SIP phone.

• On the edge router
  • NAT/QoS > Port Range Forwardin
    • voip rtp, 16384-16482, UDP, 10.0.1.10 (or whatever the addr for the VoIP adapter)

• On the VoIP adapter
  • Voice > SIP
    • NAT Support Parameters (documentation)
      • EXT RTP Port Min = 16384  (minimal RTP port used to replace private UDP port specified in SDP “m=” line)

solution 3: outbound proxy

For an example message trace refer to SIP/SDP trace using MilkFish as outbound proxy.  If incoming calls are unreliable, read this comment.

solution 4: ICE

This solution takes advantage of an external server to discover the external addr/port or to relay RTP streams.  It uses the Interactive Connectivity Establishment protocol (ICE, RFC 5245), that in turn relies on:

• Traversal Using Relays around NAT (TURN, RFC5766) to both discover external addr/port and relay messages.  When no TURN server is available it falls back to:
• Session Traversal Utilities  for NAT (STUN, RFC5389) to discover external addr/port

ICE allows the end-points (AKA agents) to discover enough information about their topologies to potentially find one or more paths by which they can communicate.  It prioritizes candidate pairs and checks the candidate pairs.  In the special case, where both end-points are behind NAT, it relies on both end-points first sending STUN requests for candidate pairs to create the binding.

Debugging and other notes

• on the SPA2102:
  • Voice > Line 1 > SIP Settings > SIP Debug Option = full
  • Voice > System > Miscellaneous Settings > Debug Server = yoursyslogserver
• or, use Wireshark
  • filter: port 5060 or udp portrange 16384-32767

For information on improving SIP routing using sipsorcery.com read these instructions.

Quality of Service (QoS) [work in progress]

VoIP phone calls are very sensitive to delay and jitter.  We are the mercy of the network for most of the route, but we do have control over our often congested edge router.  There the outgoing VoIP traffic has compete with other network traffic for limited bandwidth.  By applying simple traffic shaping we can prioritize VoIP and other interactive traffic such as SSH.

Suddenly, the old Type-of-Service (ToS) field in the IPv4 becomes more meaningful.   This paragraph describes how to use this field to label traffic types, and how to shape the outgoing traffic based on this field.  Most Linux and Windows applications use ToS for QoS.

Your DSL/Cable modem will likely relabel the QoS based on your service agreement.  Comment QoS methods for the core network are MPLS or Cisco’s Differentiated Services (DiffServ).  For more details read this overview or the DiffServ HOWTO.

Our DD-WRT edge router has priority based queuing enabled by default.  It also uses the VEGAS congestion control

netfilter rules (to be added)

…..

The priority based queuing discipline is very rudimentary.  High priority traffic will starve lower priorities.  The Linux Advanced Routing and Traffic Control HOWTO gives lots of ideas for improvements.

• Starfall

Rekenen

• Rekenweb
• Math is fun
• Primary math
• Pattern blocks
• Rekenen met SchoolTV (NL)
• Oregon Math (and other) standards

Nieuws

• Jeugdjournaal (NL)
• School TV Week Journaal (NL)
• Het weer bij ons

Science

• Yenka Technology (free after school hours), requires install
• Physics for Kids
• MIT Physics Lectures
• PhET Interactive Simulations
• Naval Electrical Engineering

Natuur

• Nieuws uit de natuur (NL)
• Huisje, boompje, beestje (NL)
• Beestenboerderij (NL)

Geschiedenis

• De middeleeuwen (NL)
• Vincent van Gogh (NL)

TV

• School TV (NL)
• Het Klokhuis (NL)

Diversen

• Email
• Boowa and Kwala

• Yenka Technology (free after school hours), requires install