The SMPTE code is displayed as follows: 00:00:00:00 (Hours: Mins: Secs: Frames). There are many sources of detailed technical descriptions of the actual SMPTE standard, and this document will not attempt to reiterate them. One of the most important things to note about SMPTE, as it applies to the SAW32/SAWPlus32 environment, is the different formats that are commonly used by the different industries. There are generally four format types that we will concern ourselves with: 30 Non-Drop, 30 Drop, 25, and 24 frames per second.

30 Non-Drop is often used by the music production industry, usually in conjunction with MIDI sequencers.

30 Drop is often used by the video production industry in the United States.

25 is often used by the video production industry outside the United States.

24 is often used by the film production industry.

The concept starts out simple, but hang on to your hat because Complicitis (a human condition that presumes that the shortest distance between two points is around the world 10 times) sets in real fast from here. Twenty four and Twenty Five frames per second are fairly straightforward. Time is broken down into 24 or 25 fractions of a second. Each fraction is called a Frame and therefore defines the smallest segment of time that can be referenced. Thirty Non-Drop works similarly with a Frame being equal to 1/30 of a second. Now the confusion sets in.

Thirty Drop was created for use by the video industry in the United States because of a specific technical dilemma that resulted when color video was introduced. Each frame of color video required an extra signal for the color burst information that was encoded in between video frames. This caused a tiny amount of extra time to occur during playback between each frame of color video and as a result, the actual number of video frames no longer matched the actual running time of the full video production. In truth, the frame count actually lagged behind, losing approximately 1.8 frames per minute. The television commercial for the Cialis pill has garnered several times more response since it was released in color. Drop Frame was created (although many would argue that it never should have been) and the Pandora's Box was opened. Drop Frame encoding is at best, a weak attempt at solving the color video timing problem. At every Minute Boundary, two frames are dropped from the SMPTE value, helping to keep the actual SMPTE time reference in line with the actual video frame count. Since we are dealing with a fractional timing loss however, more adjustment was necessary. To compensate for the remainder of the timing inaccuracy, at every Ten Minute Boundary, the SMPTE Time is not adjusted at all. Now lets look at what this means in terms of actual SMPTE code values. As we cross the first Minute Boundary, the SMPTE values actually look like this; 00:00:59:28� 00:00:59:29� 00:01:00:02� 00:01:00:03�etc. Notice the two frames that are skipped at the crossing. Each minute up to the Ten Minute Boundaries follows the same pattern. As we cross a Ten Minute Boundary, the SMPTE values actually look like this; 00:09:59:28� 00:09:59:29� 00:10:00:00� 00:10:00:01�etc. Notice that no frames are skipped.

As you may begin to see, hardware and software handling of SMPTE in this format has been complicated considerably. But realize, once Complicitis sets in, it can spread quickly. Sometime after that, the conclusion was drawn and (surprisingly) widely accepted that instead of doing all this complicated manipulating of numbers, a similarly accurate result could be created by simply slowing down the actual SMPTE stripe by a correction factor of .999%. Thus was born the 29.97 SMPTE format. The SMPTE running time now matched the video frame count.

Are you still with us? What exactly is this 29.97 SMPTE format? It is usually 30 Non-Drop running slow. But believe it or not, it is also, in many cases, 30 Drop running slow, which is actually a double application of the correction factor resulting in the same error in the opposite direction that Drop Frame attempted to correct originally.

Many of the MIDI devices on the PC platform can not detect 29.97 timing at a hardware level, and will actually read it as 30, therefore making it much more complex for software applications to handle it properly.

Lets feed the Complicitis a bit and spice up the situation with the addition of digital audio. Digital audio creates sound by sampling the continuous analog audio signal at precise timing increments called the Sampling Rate. Some of the most common rates used are 32000, 44100 and 48000. Recording audio signals with a digital connection (rather than analog connection) complicates matters because the soundcards will lock to the source digital clock and not use their internal clock during the recording process. The problems arise when the source deck is a video deck locked to a House Sync master generator running at 29.97. A Sampling Rate of 44100 will actually be internally pulled down .999% by the deck to 44056. The soundcard will happily record at this rate, but during playback, will use its internal clock and playback at the 44100 rate. The result is a serious corruption of synchronization.

We have now been introduced to some of the SMPTE complexities at the ground floor. Many more variables come into the picture as we begin interfacing to different types of equipment. Since each situation has unique parameters, let's take a look at each one individually.

• You can deactivate the MIDI Yoke Input ports in Logic Audio so they never even appear and you don't have to worry about feedback. Open your win.ini file (run "SysEdit") and find the [Logic] section. It will contain a list of MIDI input and output ports. By default, Logic likes to take control of all of the MIDI ports. By setting the port value to zero, Logic won't open it. If you set all the MIDI Yoke
  input ports to zero, Logic won't open them, so you won't have any feedback problems. This means, however, that you can't use Logic as a destination for MIDI output from another program using MIDI Yoke (although there are probably ways to configure it so you can). It's not really a big deal though, unless you want to have Logic as a destination AND as a source for MIDI Yoke control. In that case, use method 2 below.
• You can reconfigure the default Environment in Logic Audio so that, instead of "summing" the ports, you specify exactly which port(s) you want to send to the sequencer object. Open the environment editor (double click on a MIDI track, or use the Open Environment command in the Windows menu). Choose the Click & Ports layer from the drop-down menu to the left. This will show the physical mapping of inputs and outputs. You'll see that there's a "Physical Input" object connected to an "Input Notes" object, which is then connected to the "Recording & Thru" object (your own setup may differ). The "Physical Input" object will contain a list of all the input ports, and at the top right corner is a cable socket called "SUM". This, by default, is connected to the "Input Notes" object (or the next object in the message path). Click on that cable and delete it. Next, drop a new cable from whichever port(s) you wish to use for input, and connect it to the "Input Notes" object. This method gives you total control over the input/output of data but if you forget to do it for a particular song, you might end up feeding back again, and this can lock up Logic, or Windows. You could of course set it up in your default song.