-
For Walter: The cool features from EMC that have been lost in modern NLEs.
The lost cool featrues of the EMC2 (actually it’s supposed to be the EMC squared, but I forget how to make the square sign.)
This post is in response to Walter’s request for some info about some of the unique features of this long lost but pioneering NLE. For those who don’t know Editing Mahcines Corporation brought out the EMC2 editor about 3 months before the Avid came to market and holds claim as the first all digital NLE, as well as the first PC NLE. I used it during its last few years of existance in the mid 90’s.
The most substantial diffreence between the EMC and anything else was its unique way of digitizing. It didn’t use any sort of wrapper or any clip structure at all. It digitized video to a hard drive and kept a frame by frame index of what was on that drive. The index knew the reel name and timecode of every frame digitized, but there were no “clips” as such. The hard drive was simply a recorder and if you wanted to you could just go to the hard drive and start playing back as if it were a video deck. This was actually a great way to review all your material without plopping it on to a timeline.
This methodology lead to some interesting media management capabilities. When consolidating media you would be asked to specify handles, as in FCP7, but then it would also ask you to specify a length between clips, an overlap, below which it would keep them together so that if you had 2 clips that were 10 seconds apart and you specified 3 second handles with a 4 second overlap it would keep the entire sequence together, if the specified overlap was 2 seconds, you would get only the clips with the 3 second handles. This is an excellent media management feature that I wish were implemented elsewhere.
The EMC was an off-line editor; as I recall just one video track and I think 4 audio tracks, it lacked any but the most primitive EFX or CG capability. However it had some interesting features for editing in the timeline.
It had a “ripple wall” that you could temporarily place anywhere in the timeline that allowed you to ripple whatever you wanted to the left of the wall without affecting anything to the right of that divider – avoiding unwanted clip collisions while rippling.
You could take the in-point, out-point or duration of the current clip and store it in a time code cache and then use these stored values to create new points for other clips – i.e., you have a clip that you’ve “lifted” a section out of and now want to restore, take the in-point of clip 2, subtact the out-point of clip 1 and you now have a duration that you can add to the end of clip 1 to restore the full length of the clip. These cahches were great, you could use mathematical expressions to manipulate them, there were many ways to use them and I’ve longed for them ever since, to no avail.
When copying and pasting in overwrite mode the older clips were covered over but not replaced unless you specified. You could reverse the stacking order of the top clips to reveal the underlying clips, and then go back again, giving you a somehwat limited version of the stacking feature in DAWs. It wasn’t a very sophisticated feature, it got confusing as to which clips, old or new you were working on, but it was there and could be used until you merged down and eliminated the underlying clips.
It had a very sophisticated, for the time, multi-cam feature. You could only see the multicam window when the timeline was paused, but you could cut as you played. The multicam would automatically sync up any clips that had the same timecode as the timeline clip, you could specify that only certain reel #s were to be looked at, you could add time code offsets to any or all of the various reels. It was quite the fastest setup for any multicam I ever worked with and as good as anything that existed at the time.
At the click of a button you could switch from timeline view to EDL view and see the CMX style EDL of the timeline you were working on. You could edit and make changes in the EDL and those changes would be reflected in the timeline when you switched back. While I didn’t use this a ton, if you know how to read an EDL it can be quite handy at times. I don’tknow that it would have that much application now, with the enormous number of video and audio tracks in a modern timeline, but some sort of index view can often be valuable when looking for certain metadata and timecode relationships that a graphical timeline often hides.
The EMC was bought up by Dynatech in the mid 90s and they EOL’d it a few years later. I think Bill Ferster, the original inventor, still teaches at the University of Virginia.
Herb Sevush
Zebra Productions
—————————
nothin’ attached to nothin’
“Deciding the spine is the process of editing” F. Bieberkopf -
5 Replies
-
Herb, this was a fascinating read. Many thanks for sharing!
The sad part is seeing how some of these great features are linked to ways of working that we as an industry have been pushing away from — transcode on import, offline/online workflows, caring if timecode even exists at all…
I suppose these are all two-edged swords, with notable advantages and disadvantages.
Seeing the number of requests for Premiere to adopt a native format does give me hope that not all these things will be lost.
[Herb Sevush] “The most substantial diffreence between the EMC and anything else was its unique way of digitizing. It didn’t use any sort of wrapper or any clip structure at all. It digitized video to a hard drive and kept a frame by frame index of what was on that drive. The index knew the reel name and timecode of every frame digitized, but there were no “clips” as such.”
This sounds much the same way that Smoke (pre-2013) preferred to work; it had a framestore, and all hard imports were transcoded to DPX image sequences upon import. This legacy architecture is still around in Smoke 2013 now, even as it adds native media support.
As a graphics guy, I love image sequences, and I think that some video editors don’t know what they’re missing. Disk requirements are high, but since you’re generally using uncompressed formats, decode complexity is low and generational loss is minimal or nonexistent. All the major image formats are perfectly cross-platform, including not just Windows and Mac but also Linux. Smart renders are a simple file copy instead of some arcane QuickTime dark magic. Many command-line image processing tools work well with image sequences.
Avid and FCPX don’t have framestorse, but they do have managed media stores, so they preserve some (but not all) of these advantages. Premiere seems to be the furthest one out here, working solely with unmanaged media, though it does have pretty good image sequence support.
[Herb Sevush] “When consolidating media you would be asked to specify handles, as in FCP7, but then it would also ask you to specify a length between clips, an overlap, below which it would keep them together so that if you had 2 clips that were 10 seconds apart and you specified 3 second handles with a 4 second overlap it would keep the entire sequence together, if the specified overlap was 2 seconds, you would get only the clips with the 3 second handles. This is an excellent media management feature that I wish were implemented elsewhere.”
Wow. That is so much smarter than slightly separated or occasionally slightly overlapped consolidated media.
Does anyone here know how Avid handles consolidation like this?
Pardon my ignorance, but can FCPX trim when it consolidates? How does it work?
[Herb Sevush] “It had a “ripple wall” that you could temporarily place anywhere in the timeline that allowed you to ripple whatever you wanted to the left of the wall without affecting anything to the right of that divider – avoiding unwanted clip collisions while rippling.”
I don’t just want this in my NLE, I want this in every time-based application I use. This would have broad application in keyframe management for motion graphics and 3D animation as well the layer-based media servers I use.
[Herb Sevush] “When copying and pasting in overwrite mode the older clips were covered over but not replaced unless you specified. You could reverse the stacking order of the top clips to reveal the underlying clips, and then go back again, giving you a somehwat limited version of the stacking feature in DAWs. It wasn’t a very sophisticated feature, it got confusing as to which clips, old or new you were working on, but it was there and could be used until you merged down and eliminated the underlying clips.”
Non-destructive overwrite on collision.
I suppose FCP Legend’s behavior keeps the timeline perfectly consistent with the canvas output, but irretrievably destroying user data is not a good thing.
Dataton WATCHOUT (a real-time, networked, multiple-output media server/compositor) does something a bit like this, too. If two pieces of media end up on the same layer at the same time, it highlights the overlapped frames with a little red line. WATCHOUT doesn’t have a stacking system to deal with it — it’s up to the designer to rearrange the clips to eliminate the conflict — but unlike FCP Legend, this sort of reaction to a timeline maneuver destroys no user data.
[Herb Sevush] “At the click of a button you could switch from timeline view to EDL view and see the CMX style EDL of the timeline you were working on. You could edit and make changes in the EDL and those changes would be reflected in the timeline when you switched back. While I didn’t use this a ton, if you know how to read an EDL it can be quite handy at times. I don’tknow that it would have that much application now, with the enormous number of video and audio tracks in a modern timeline, but some sort of index view can often be valuable when looking for certain metadata and timecode relationships that a graphical timeline often hides.”
Score one for FCPX. Its timeline index is very cool, though it’s only a view of the timeline data, not a two-way view/editor like EMC^2’s.
[Herb Sevush] “I think Bill Ferster, the original inventor, still teaches at the University of Virginia.”
Now we just need to get him to join CreativeCOW…
Again, Herb, thanks for writing this up.
Walter Soyka
Principal & Designer at Keen Live
Motion Graphics, Widescreen Events, Presentation Design, and Consulting
RenderBreak Blog – What I’m thinking when my workstation’s thinking
Creative Cow Forum Host: Live & Stage Events -
[Walter Soyka] “Again, Herb, thanks for writing this up.”
One of the fun things about working with EMC back then was that if you gave a suggestion for a new feature, or spotted a bug, when the next upgrade came out they gave credit to the user where applicable – the one time I saw my name in the release notes it was better than winning an Emmy. Talk about building brand support.
Herb Sevush
Zebra Productions
—————————
nothin’ attached to nothin’
“Deciding the spine is the process of editing” F. Bieberkopf -
Very late joining this thread, and new to the forum (though I’ve lurked for years) –
I was one of the early proponents of EMC in the networks in New York. I think our team at CBS News Productions was the first to edit a whole series and output directly for air from the machines. It was a series called 20th Century with Mike Wallace, for A&E. Loved that machine. Forced to work on Avid for much of the rest of my career but EMC was the first with the best as far as I was concerned. I was still outputting direct to tape for air (and feeding the satellite live!) from the EMC at ABC News Productions ’96-’99 for Discovery News.
Now I’m designing a new community science and media center (just built a planetarium!) and ironically having the FCPX vs legacy FCP or Avid debate with our junior video teachers. I try to tell them they have to learn FCPX and teach it because we can’t go back. The 20-somethings already want to hold onto what they know. But that is a terrible thing to teach young kids coming up – they have to know that they will be learning new technology at an accelerating pace all their lives. So it’s the leap to new Mac Pros and FCPX and 4K – why can’t they see that?
Thanks for the reminder of the early days – pioneers then, pioneers now.
-
[Dave Pentecost] “I was one of the early proponents of EMC in the networks in New York.”
Always good to hear from a fellow EMC’r. Most editors falsely assume that Avid was the first digital NLE. I always thought that the EMC was the betamax of NLE’s: first out, best quality, but out marketed and soon forgotten.
[Dave Pentecost] “I think our team at CBS News Productions was the first to edit a whole series and output directly for air from the machines.”
Interesting. I never used the EMC for air, I didn’t think the video quality was good enough, although I did use the audio output in a fairly unique workflow that I came up with. You didn’t mind the video artifacting and the lack of any real CG?
Herb Sevush
Zebra Productions
—————————
nothin’ attached to nothin’
“Deciding the spine is the process of editing” F. Bieberkopf -
For the time, the artifacts were minimal. We were producing historical documentaries from primarily CBS archival film transfers. They started good and stayed acceptable through a final post with host wraps and bumpers. The later series at ABC (weekly science news show, original reporting) was composited with all bumpers, teases and studio intros in the EMC and fed by satellite to Discovery headquarters (sometimes directly to the network if we were down to the wire). As in everything those days, we just had faith we would get through, and we always did – no equipment issues or freezes in 200 shows. I don’t miss the stress now, but it was an exciting time.
Reply to this Discussion! Login or Sign Up