Nubies' Corner

     Digital Audio can be one of the most confusing operations on a Mac, but with the right set of external speakers, a sub-woofer, and a basic understanding, the sky is the limit. If you can do it in analog, you can do it in digital, and then some.

     (MIDI and other assembly formats such as MOD, S3M, and others will be discussed in a later issue).

     There are four primary factors to an audio codec. A codec is a master print of of the way an audio waveform is stored digitally. The format most encounter is "straight audio" on CD's. Straight audio pretty much says what it is, a direct time segment representation of the audio waveform. In each segment, the waveform has an entry value and an exit value (the entry value of the next segment). Hence, each segment is defined by its single digital value. When represented on a graph, a straight line connects the points.

     As the time point shrink (which is hardware limited, check your specs, though in most uses, the hardware is more than up to the task), the straight lines between points become shorter and more frequent. The more the points are defined, the closer to the the original analog waveform the pattern represents.

     44KHz is the industry standard clock time for time segment definition, which is called "sampling rate". 22KHz is not as crisp, but it takes less memory, and occupies the high end of the onboard sound systems. The low end is the even more economical 11KHz sampling rate. Even lower rates are used for sound transfer on the net, such as Netscape's module AIM, RealAudio, and some others who want to stream audio unbroken content over the net.

     In essence, one second of uncompressed straight audio has 44,100 bytes of information that pertain only to the waveform. One minute of audio contains 2,646,000 bytes of information per track. Stereo, 5,292,000 bytes per minute.

     As the time base can vary the segment length of the horizontal axis of the graph, the number of bits per byte can change the segment length in the vertical axis, making more value choices available. As a rule, the number of bytes come in 8 bit chunks. 8 bit, 16 bit, and 32 bit sound are the most common. CD audio uses 16 bit. This means that one minute of CD stereo audio uses 10,584,000 bytes of information (10 meg a minute), and surround sound tracks use about 20 M/min. There is also some management data moving along with it, so the stereo CD runs about 12 meg a minute or 200K per second.

     And third, mentioned above is the number of tracks. 32 track digital recorders eat memory for breakfast and as between meal snacks before they even get warmed up. As an obvious rule, the fewer the tracks, the less memory that is required. A decent recording studio should have a minimum of 10 gig as the staging volume for the mastering duties alone. Then there is storage............. Pro audio requires a lot of drive space and specialized hardware.

     Fourth, the encoding/decoding schemes (mathematical formulas for processing and storing data and making the file size smaller) has become the big race on the net. Everyone is looking to be the top dog in the compression race that the MP3 quagmire currently holds. MP3 is not the best, it's only the most popular because the developers of WinAmp got lucky, and everyone else jumped on the obvious bandwagon simply because it was the first there. There are better codecs out there that use less than a fourth of the size for comparable sound. They're just not the subject of a lemming-like rush to their advantages. There are many codecs out there just dying to be used, and the choice is not an easy one to make when encoding. They use a wide variety of approaches, and some work better than others. Study, record, and listen like crazy to the differences.

      So, the best popular sound format that is universally playable is CD sound at 44KHz, 16 bit, stereo, uncompressed. A CD can hold a little over an hour of music, about 20 radio segment length songs. The most common straight audio on the net is WAVE (Windows native audio, .wav) at 11KHz, 8 bit, mono. Real Audio has a usable 5.6KHz, 8 bit, mono, compressed format that has set many standards for small package size.

      However, the best audio package that I've seen is QuickTime 4 audio using QDesign Pro 2 codec compression that delivers 11KHz, 16 bit, mono sound in less space than the Real Audio, and it sounds very good by net standards. At 44KHz, 16 bit, stereo, it kicks MP3 all over the place with far smaller packages. You need to go no further than the native Mac QuickTime Player to get the best out of it. A QDesign Pro 2 encoded QT movie (audio track only) can rip a 40M song from a CD to less than 1.5M and present it faithfully in full 44KHz, 16 bit, stereo sound that any platform can play using QT. The codec isn't cheap, but it's the best out there, and people are paying notice. It's only needed to encode, not decode, as QT handles that without any enhancement. Many new web sites are going QT and QDesign for net streaming and doing a great job of it, getting rave reviews by those who are set up with QT 4 to stream them. This is the format I use for my own onboard library.

     The vast array of audio formats out there (a few hundred audio formats), which can be extremely dizzying at times, can be pretty much summed up in the four aspects: sample rate, byte size, tracks, and compression. Finding the right ones takes a lot of testing, which develops the skills and knowledge that the true digital artist must acquire along the way to earn full credentials. Grabbing a few free all purpose players like QuickTime, PlayerPro, and Sound App can make short order of being able to play most of whatever comes your Mac's way. Grab a bunch of the free and shareware players and find your own favorite. As far as making it.......
     The rest is iniative and learning curve and $$$$$$. But it sure sounds sweet when you get it right.






Software

     Poser Pro Pack was recently released by Curious Labs, inc. to augment Poser 4 to the next level.

     The installation overwrites the original Poser application and Runtime folder. Making a back up of the original folder (especially the Runtime folder) is highly encouraged, since many of the links to obj, textures, and other attributes could well be negated. It takes a bit of following up, checking all the saved .pz3 files to see what links were damaged in the rewrite. Double checking with the installation log is the only way to determine if the library files need replacing or the scene files needs reassigning. Expect a transition in regard to old work in Poser. Hence, it's best to install it right away if possible.

     At first glance, the half dozen Daz cartoon characters are the first thing to grab the attention. I can tell that they will add themselves to the free cartoon kids that Daz has on their site for download. They are great characters and will no doubt see plenty of use in advertising and on the net. But the changes are far deeper than the cute new cartoon village.

     The second item of notice is the second interface, tabbed behind the familiar Poser 4 interface. This is the animation workshop. Center and foremost is the bone lab, whose window allows for assignments of parameters to the figure's digital skeleton. The lab can set limits to the way a limb bends to facilitate motion sequences, cutting down on the possible processing errors. The depth of the lab is a matter for weeks of study to absorb the nuances built into it.
     Features are plentiful, including IK chains and user default settings, rotation orders, joint parameter assignment, and lots of bone assignment and testing tools. This is also a great area for creating animation ready figures from scratch, using imported .obj model parts.

     Then there is the Python interface, which brings cross-platform scripting to Poser. AppleScript might have been easier to use, but Poser is a multi-platform applications with parts that can be shared across platforms. Python works on any machine and is growing as a cross-platform scripting tool. To use Python effectively, MacPython is needed as an editor (or BBEdit, Alpha, or whatever your favorite code writing WP happens to be.) MacPython 2.0 (freeware), like Python for the other platforms, can be found here. It does have advantages when dealing with Python.
     One of the major problems that I found with Poser was having to set ambient light levels all to the same point to avoid the default "moon shadows" that drove me crazy with those harsh shadow lines in the skin textures at default. In the Poser 4 interface, ambience is set for every single mesh in the material labs. Some simple assemblies can have hundreds of objects. Imagine setting the ambient level for every single pearl in a long pearl necklace (posable prop), each single pearl needing to be set for ambience, reflection, highlight, and base color. Tedium doesn't begin to describe the process. Metacreation's greatest fault was not putting a select all objects command in the materials lab, for base universal scene settings.
     But now Python can be used in its incorporated interface to run scripts that can be set manually. (It's easier than POV-Ray, so I'm cheering loudly.) The scripts can be manually set from a template, simply entering the values for the group on which the object(s) require in the scene, and all that is done, saving a lot of time and thankless set up time and effort. Until Poser reaches a more sophisticated level in its editors, allowing better group commands, this Python scripting will be invaluable for setting complex defaults that save the majority of the effort that is required to bring the scenes under control.

     Another new feature is the view selection buttons on the side of the image frame that allow for the traditional 4 way frames, front, top, side, selected camera simultaneous views, as many of the other 3D modelers have. It also lists the camera providing the view in the frame, something that annoys me, since I have not been able to remove it from the frame in user default settings. It needs to be turned off manually with every new frame. (Something for the next update to fix, as that should be an option, not a mandatory feature in the default.) A lot of these new features are making a stab at bringing Poser into the same basic interface as the rest of the 3D modelers without surrendering the trademark Metacreation's interface that runs through RayDream, Carrara, Bryce, Poser, etc. In many ways, it makes it more friendly to cross-application work.

     Exporting to new formats is one of the biggies. Motion export to 3DS and Lightwave, as well as Flash, has made Poser more of a mainstream digital animation tool, trying to take a crack at some of the mainframe applications that have fueled most of digital animation thus far. For the animator, the Lightwave (5.X and 6.x) plug-in that allows it to accept Poser imports essentially throws Poser into the TV and movie animation business. Same with 3DS. But of more importance to web designers, it exports Flash .fla, which can be further edited in Flash (much as Photoshop is used for post production in regular Poser renderings). The display settings are critical for a good flash export, as well as a good estimate of the number of colors needed to make an acceptable vectorization. Nonetheless, this means a lot more 3D character animation on the web in the coming future.

     Lots of little effects thrown into the upgrade make it even better as a rendering engine.
     Motion blur is the most notable, to make it more compatible with film based cinematography.
     Material animation is another new trick with a lot of possibilities, but the interface is still hopelessly manual oriented. Scripting will be required to use this feature efficiently.
     Animation sets have been added to augment the Viewpoint Media Player (the new version of Metastream).
     File compression is another feature (control in preferences) that allows for smaller .pz3 and .cr2 files, which are labeled .pzz and .crz, which can reduce the rather sizable file size while keeping them ready for use.

     I will note that there are some differences between Poser 4 files and Poser Pro Pack files that make them incompatible in certain cases. It might be a wise practice to duplicate the full Poser installation and update that, leaving the working copy of Poser intact. Then as needed, items outfitted in Poser 4 could be copied over to the Pro Pack folder for further modifications and animation. Above all, save an extra copy of the original RunTime folder, from which the original Poser set up can be reproduced with a fresh installation of Poser. Running dual versions can solve a number of problems, especially if one is using net acquired models and other Runtime files.






Software

     TypeStyler is an essential Mac design tool for creating font graphics.

Kiss
© 1999

     Long one of my favorite graphic tools for most of my decorative text, TypeStyler takes system installed fonts and decorates them with zone, shadow, emboss, chisel, and many other effects, including the use of a custom user image library. Images can be copied into the editor and used in the overlay process, allowing for endless graphic variations.
     The interface is highly compartmentalized, with a variety of windows that can be opened as needed. The Main window contains the current text and applied effects. The display works vectorially, so resizing on the fly is smooth and easy. The window can have set margins and be assembled as a web page or printable letter, or it can be used for its parts and exported to be discreet elements in a wide variety of formats, from JPEG and GIF to Photoshop and TIFF and more. The main window is a very flexible palette that holds all the necessary tools to work efficiently and certainly. With enough RAM, it can be opened side by side with Photoshop for sizing purposes, which is convenient with its transparent background Photoshop layer exports.
     The tool bar is a little confusing at first for most Photoshop users, as their function is usually other than what is expected. But once the user learns the ropes with the tools, it's a very intuitive set up. The hardest learning what needs to be set from the menus for the tools to work properly. But the work habits are easy enough to remember, as they are logical. Essentially, there is a text cursor that nails a spot on the canvas. This spot then opens the Type Attributes window.
     This is the word processing interface where the text is edited and the other function windows can be accessed.

 

     When set and the okay button is pressed, the text is constructed on the main window.

 

     Font dynamics, line kerning, line spacing, writing order (for overlap) in the Text Attribute window.

 

     Bezier controls are available for making top and bottom paths (up to three spots).

     These are the main shaping controls. The other side of the application is the Style Workshop and Style Library, where the motif is set and stored as presets.

 

      This is where the coloration, highlighting, shadowing, expanding, pillowing, and otherwise adding some visual snap to plain font text. This is where the real impact is found. Use of images atop the font is a powerful tool, and the Custom Image Library allows users to supply their own images for use in the styling.

 

     These are the main tools, but there are so many more, like the fill editor, wacky shape editor, Javascript rollerover editor, GIF animator, caps editor, and on and on and on. Combined with a hefty copy of Photoshop as an image editor, virtually any motif can be set to editable text and then set in any image. TypeStyler expands the capacity of Photoshop well beyond its native capacity. If you want to see examples, check my sci-fi title archive.
     It exports transparent GIF's directly to file, compressed and ready for posting on the net. Even without Photoshop, it's an invaluable tool. It imports quite well into Flash, especially if the text is color aware.