CMX Model 600
 
Home
 
Honor Roll
 
Ampex
CDL
CMX
Convergence
Datatron
EECO
EPIC
ISC
Mach One
RCA
Smith
 
EDL's
Offline
PC Editors
Specials
Sweetening
Telecine
 
HS100/200
TRT-1B
VR-1000
 
Credits
 
 
Links
Japanese
 
 
Picture of the CMX 600 Console

 

 
In the Beginning...
CMX was formed as a joint venture between CBS and Memorex. Their goal was to produce a revolutionary new offline video editing system. This system stored the video on computer disks, which provided instant access to both picture and sound, and the ability to make, and see, changes in real time.
 
This new systems was called a Random Access Video Editor (R.A.V.E). Adrian Ettlinger, of CBS, provided the concept for the new system, and Memorex, then heavily involved in data recording, provided the disk recording technology.
 

 
The Editors View...
Picture of light Pen Complete control of the editing process was done with a Light Pen. The editing console (pictured above) housed two black and white monitors. The monitor on the right was used by the editor to select scenes, select cut points, and to control playback. All this was done by clicking on parts of the screen.
 
The left hand monitor showed the edited sequence. You could at any time play the sequence, forward, or backward, and at a variety of speeds.
 

 
The Back Room...
As simple as the control console was, the equipment necessary to make the 600 preform its magic was substantial.
 
The CMX-600 consisted of, an ASR-33 Teletype which was used to control the recording process, and to input, and, output the Edit Decision list.
 
The electronics for the 600 filled two equipment racks. The first held the monitoring equipment, PDP-11 computer, with its 16k of memory, and interface electronics. The second held the video and audio signal systems, and most important, the "Skip Field Recorder" (more about this in a moment). And finally, the Disk drives.
 
The CMX-600 could support up to six (6) disk drives. Each drive was about the size of a washing machine, and required 220v 3-phase power. With six drives, the system could store 27-minutes of audio and video.
 
In order to maximize the recording time, only one field of each TV frame was recorded. Using the Skip Field Recorder, audio from both fields were combined with the single field of video, and recorded on the disk drives. Each track on the disk held one video field with audio from both fields. The recording was analog, and the audio was stored in the video's horizontal interval.
 
During playback the Skip Field Recorder was used to artificially create an interlaced picture, by first playing back the field without delay, then playing it back with a one-half line delay.
 
Picture of disk pack Each disk drive held one disk pack. These disk packs have 11 platters (20 recording surfaces), each platter is 14" in diameter. The disk pack holds a total of 39-megabytes of data.
 


CMX-600 Demo
In 1971 a promotional video of the CMX-600 was produced at CBS Hollywood.

Click here to view the 13 minute video (requires Microsoft Windows Media Player).
 


 
Memories...

From Dave Bargen [updated below]:

The following are a few of my memories of the making of the CMX 600 Demo tape (Now DVD)...they may be close to correct...

CMX, the company, was started in April 1970, as a joint venture of CBS and Memorex. The goal was to develop and market the first electronic random-access editing system. CBS wanted this to speed the transition from film to video tape. Memorex wanted this to develop a new market for their disk memory systems, and to enhance their video tape sales.

I joined the company in May of 1970 as Systems Engineer, responsible for the general structure of the system, and the computer control. By early March of 1971, the 600 was functional...an impressive achievement for taking a new concept to working hardware in 10 months. In March of 1971, the Demo was started. The raw material was shot in LA, probably on a set at Studio City. Then a CBS crew came up to CMX in Sunnyvale, and shot the material showing editing on the console on quad tape. The editor was Harry Black (or White?). As can be seen in a close-up, he was quite nervous...because this was the first time he had worked with the system, and because there was a lot of jitter in the light pen control.

The demo tape implies that it was finished automatically on tape, based on the editing on the 600. That was the concept, but not the reality. The reason is that the assembler portion of the system, the CMX 400, which was to control the quad tapes, had not yet been built.

So the demo tape was actually made back in LA, after the shoot in Sunnyvale. I am not sure if it was edited on tape using conventional techniques (Editec or EECO), or edited on film, or a combination of the techniques.

In any event, the Demo ended up on Quad tape and was taken to NAB in April of 1971, which was in Washington DC that year. The tape was shown repeatedly in the Memorex booth, and drew a very positive response.

The Assembler portion of the system was functional in early summer of 1971, and several systems were shipped in the summer and fall of that year.

Before the CMX 600, CBS had an in-house project in New York. Adrian Ettlinger led the technical work. The system simulated random-access in a very limited way, using multiple Sony 2-inch (!) helical-scan tape machines, and a PDP-8 computer.

And before that was film strips, hanging on hooks.

Computer:

  • DEC PDP -11
  • 16,000 words of core memory (16 bit), or 32 k-bytes
  • Blazing speed: 3 to 5 microseconds per instruction. (About 1000 times slower than a typical PC today.)
  • No hard drive
Disk Pack Video Storage:
  • Memorex digital disk drives
  • Each about the size of a washing machine
  • Drives modified to store and retrieve analog video signals
  • Capacity: about 5 min video per disk pack
  • Usually 4 to 6 drives per system, or 20 to 30 minutes of storage per system
Video System:
  • Bandwidth of the disk drives was less than 1 MHz...hence black and white video, without compression
  • Every other field of video was recorded on the disk packs, to extend storage capacity
  • The "lost" field was recreated using a 1-field disk memory in the 600 main unit
Audio Encoding:
  • Encoded as PCM data on the back porch of the video horizontal sync signal
  • Plus: Could hear audio in still-frame, slow-mo, etc
  • Minus: Quality poor, resulting from sampling problems due to jitter in the signal from the disk drives.
Light Pen Control:
  • This was a technology used primarily by the military
  • CBS wanted to avoid conventional controls, to minimize union jurisdiction issues
  • It also provided some "gee-whiz" appeal
  • The jitter again was due to jitter in the signal form the disk drives
Software Menus/Controls
The names of the controls on the screens and the logic of the operations were based on conversations with film editors, and observation of them cutting film. The goal was to make it as easy as possible for a film editor to operate the new system.
Console:
A mockup of the console was first built out of plywood. It had adjustable tilt and height of the desk surface, monitors, etc. This was used to arrive at the desired "human engineering" parameters. Then an outside design firm was hired to make the final model.
CMX People
Bill Butler - President
Gerry Heitel - Sales Manager
Jerry Youngstrom - Engineering Manager
Dave Bargen - System Engineering Manager
Jim Adams - Programmer
Steve Foreman - Programmer
Will Pearson - Video Engineer
Yves Faroudja - Contract video engineer
Gene Simon - Super Tech
CBS People
Joe Flaherty - VP R&D
Bill ? - Liaison - CBS/CMX
David Horrowitz - Liasion - CBS/CMX
Awards:
  • CMX received an Emmy award for the 600
  • The design group received a design award for the console


Updated 01/30/2007:

Here's a little clarification of the early 600 time-line, as best as I know it:

CMX, as a company, was formed in the spring of 1970, as a joint venture between CBS and Memorex. (It was changed to a corporation in the spring of 1971.) However, some work on the ideas of CMX preceded those dates:

Work on the idea of a random-access editor started at CBS, perhaps a year or two earlier, resulting in the tape-based RAVE system, as a demonstration of concept. Adrian Ettlinger would know more about the dates.

Work on the idea and techniques of random-access storage of video on computer disk drives was started at Memorex before CMX was formed, though I'm not clear on when the exploratory work began. Note that several of the key players of CMX had come to Memorex from Ampex, with their experience of magnetic video recording.

It was the merging of these parallel development ideas which resulted in the formation of CMX.


From Steve Foreman:

I joined CMX in March of 1971 at the peak of the CMX-600 development when CMX was finishing the CMX-600 engineering unit and demonstrating it with Howard Smith (CBS Editor) in Sunnyvale.

The CMX-600 disks were modified Memorex units modified to operate at a lower rotational speed to and hence a lot of heads died during the development. My first programming task on the PDP-11 was to devise a table to quickly remove a disk head from the software address calculation. The NTSC disks capacity was 5.4 minutes and PAL was 4.5 minutes video material(B/W). The software was loaded via an asr-33. This was a very slow process. During my first week at CMX I had to observe the demonstrations from outside the editing suite and if the process hung, I was to reset the computer. This was a crucial part of the demonstration as the design concept was to have a minimum of buttons, knobs etc on the editing console so the editor would only be concerned with artistic activity.

In the summer of 1971 the CMX-600 was set-up at the Republic Studios (Burbank?) on the main auditorium stage for demonstration to the board of directors to (I think) continue funding or to demonstrate what progress had been achieved. Dave, Jim, Bill et al and I attended.

Following that demonstration the CMX-600 was set-up in an office at the site where production tests/work was conducted. A gentleman pictured in the demo DVD you sent me (can't remember his name), grey headed, glasses, walked with a slight limp managed the work. He had worked in New York at CBS centre.

My work then (summer 1971) was related to making fixes to the software and visiting the site several times. I think the first customer outside California was EUE Screen Gems in New York. Then Rank studios in London was the first/only(?) overseas site. I had many long telephone conversations with Davey Morgenstien who worked in London.

During the time the CMX-600 was working at the at the Republic site, there was an industrial dispute between the film editors and video editors. The dispute was due to the CMX-600 being an electronic instrument but employing film editing techniques.

I remained with CMX through the Bill Orr purchase and prior to the move from Sunnyvale. I remember Bill called together the company and presented a cheque for five million to a CBS/Memorex representative. I left Sunnyvale for travel around the US/Canada during mid-summer 1975 and subsequent departure to Australia. I had some contact with CMX whilst here. I designed and patched the CMX-300 for Colorfilm P/L to enable automated A-B rolls editing. Once the edit was started the s/w automatically switched the video switcher between the "A" and "B" rolls(tape) with pre-programmed effects and transition durations.

I trust this is an accurate account of events during that time. Not responsible for failing memory due to the aging process.

 

Back to Offline