Amiga.org
Amiga News and Community Announcements => Amiga News and Community Announcements => Topic started by: Pyromania on August 09, 2014, 04:25:34 PM
-
Nice article about the original Video Toaster, also gives the Amiga a nice mention.
"Television is a complicated, expensive business. Or at least it used to be : in the 1980s, it cost a small fortune to get the equipment to shoot and broadcast your own TV show that looked as good as the big networks. However, one gadget changed all that, making it possible to produce good-looking TV shows with multiple cameras, titles and special effects without a big network budget. By doing this, our gadget set the stage for the cable TV and Internet broadcasting revolution: the Video Toaster."
Read the rest via the link below.
https://medium.com/people-gadgets/the-gadget-we-miss-the-video-toaster-93509e978549
-
Thanks for that link :)
This answer my question about the NTSC signal, but it looks like it's possible to use it in Europe!
I need to find the Amiga Format magazine issue 52 ;)
-
Nice article about the original Video Toaster, also gives the Amiga a nice mention.
"Television is a complicated, expensive business. Or at least it used to be : in the 1980s, it cost a small fortune to get the equipment to shoot and broadcast your own TV show that looked as good as the big networks. However, one gadget changed all that, making it possible to produce good-looking TV shows with multiple cameras, titles and special effects without a big network budget. By doing this, our gadget set the stage for the cable TV and Internet broadcasting revolution: the Video Toaster."
Read the rest via the link below.
https://medium.com/people-gadgets/the-gadget-we-miss-the-video-toaster-93509e978549
Wow what a brilliant post this is and a awesome video toaster which I never knew could do so very much and top quality video, now where can I buy one from? really enjoyed this my friend very best wishes Brian.
-
Commodore really screwed the pooch here. This was, finally, their "killer app", one that compels people, often irrationally, to buy it and buy whatever computer is needed to run it. Like so many other "killer apps" you have a lot of purchases being made by people who have no real use for it, who won't really use it much at all much less to its fullest, but this is ultimately the goal and a practically necessary component for the economic health of media-based developers.
You need a large number of people with no business buying your product to, in fact, buy your product so that the people who really use it can enjoy it at a lower cost and with the security that their needs will be supported long term. This has a "trickle down" effect with related and even non-related titles. Folks buying Amigas with the idea they might buy a Toaster, folks buying other software to do other things on the Amiga+Toaster they just bought, etc.
-
Wow what a brilliant post this is and a awesome video toaster which I never knew could do so very much and top quality video, now where can I buy one from? really enjoyed this my friend very best wishes Brian.
Being an Amiga user who started using them around 1986, or early 1987, and also since I live in the USA, I take for granted that most Amiga users are well acquainted with NewTek's Video Toaster & Flyer products, until I come across forum messages like yours above. I also forget that not all Amiga users have been users as long as myself, and many of them became interested in the Amiga in very different ways than myself, so they may not have been exposed or interested in video products.
The Video Toaster is what made the Amiga famous in North America, where we did not have nearly the acceptance and popularity of Amiga computers that was present in Europe. I cannot understand why NewTek never produced a PAL compatible Video Toaster board, as they surely would have sold hundreds of thousands of them, unless they looked at the sales numbers (if available anywhere), and saw that the most popular Amiga models in Europe were the A500 and later the A1200, which could not use a Toaster board without converting them to tower cases and adding video slots.
Maybe NewTek was making so much money in North America at the time, they did not worry about making a PAL version, or maybe they were already thinking of moving away from the Amiga due to Commodore's troubles, and they did not want to spend any development resources on anything except moving to the PC. I have never seen a definite answer why NewTek did not make a PAL compatible Toaster board.
I have several Video Toaster/Flyer systems and know a personal friend that I went to High School with who made his living with the Toaster for many years, before the NTSC video standard was replaced by digital HD video standards. The Video Toaster & Flyer is still a great Amiga tool to play around with, and can still be quite useful for video projects that don't require resolutions greater than the NTSC format.
As for buying a Video Toaster for use in the UK, I am not sure it is worth the trouble of also needing to convert all video signals to or from the PAL format to the NTSC format. Others that have already done it will have to advise you on what is needed and how good the results are.
If you can't find a Video Toaster board to purchase in the UK at a decent price, let me know, as I can probably afford to sell you one of my backup boards. Good luck and if you do get a Toaster, I am sure you will enjoy hundreds of hours of fun with it.:)
-
Being an Amiga user who started using them around 1986, or early 1987, and also since I live in the USA, I take for granted that most Amiga users are well acquainted with NewTek's Video Toaster & Flyer products, until I come across forum messages like yours above. I also forget that not all Amiga users have been users as long as myself, and many of them became interested in the Amiga in very different ways than myself, so they may not have been exposed or interested in video products.
The Video Toaster is what made the Amiga famous in North America, where we did not have nearly the acceptance and popularity of Amiga computers that was present in Europe. I cannot understand why NewTek never produced a PAL compatible Video Toaster board, as they surely would have sold hundreds of thousands of them, unless they looked at the sales numbers (if available anywhere), and saw that the most popular Amiga models in Europe were the A500 and later the A1200, which could not use a Toaster board without converting them to tower cases and adding video slots.
Maybe NewTek was making so much money in North America at the time, they did not worry about making a PAL version, or maybe they were already thinking of moving away from the Amiga due to Commodore's troubles, and they did not want to spend any development resources on anything except moving to the PC. I have never seen a definite answer why NewTek did not make a PAL compatible Toaster board.
I have several Video Toaster/Flyer systems and know a personal friend that I went to High School with who made his living with the Toaster for many years, before the NTSC video standard was replaced by digital HD video standards. The Video Toaster & Flyer is still a great Amiga tool to play around with, and can still be quite useful for video projects that don't require resolutions greater than the NTSC format.
As for buying a Video Toaster for use in the UK, I am not sure it is worth the trouble of also needing to convert all video signals to or from the PAL format to the NTSC format. Others that have already done it will have to advise you on what is needed and how good the results are.
If you can't find a Video Toaster board to purchase in the UK at a decent price, let me know, as I can probably afford to sell you one of my backup boards. Good luck and if you do get a Toaster, I am sure you will enjoy hundreds of hours of fun with it.:)
Hi Amigadave
thanks for all the information my friend and you are right some of us Amiga users are relatively new to the Amiga and its only in the last few years I became caught by the Amiga bug and I am still learning and the video toaster amazed me so I am reading up on them but thanks for the offer to buy from you I really appreciate the offer, I have not looked for one in the uk yet but I have seen them come up occasionally so its a little more reading before I take the plunge but still fascinating reading,very best wishes Brian.
-
...
Maybe NewTek was making so much money in North America at the time, they did not worry about making a PAL version, or maybe they were already thinking of moving away from the Amiga due to Commodore's troubles, and they did not want to spend any development resources on anything except moving to the PC. I have never seen a definite answer why NewTek did not make a PAL compatible Toaster board...
Well, they were making money hand over fist. I recall a picture of the two founder's matching Ferrari and private jets. They were a pretty big deal in Kansas before they moved. The margins must have been insane on those things. They were living like rock stars. It was pretty easy to see how the money changed Allen Hastings too.
Given that the Amiga host processor didn't do any heavy lifting at all, no video data passed into the Amiga, the DVE effects were all done by custom chips on the actual VT board, I still don't understand why the folklore at least was that it was intrinsically tied to anything that the Amiga was actually doing besides supplying that NTSC compatible clock signal.
Seems like making a PAL version or a version that worked in a PC/Mac with an onboard sync generator as an additional component would have been a no brainer. TruVision had been making pro-level NTSC gear for PC and Mac for years, just not with realtime DVE and switching, which was Grass Valley territory. NewTek was just loyal, I guess. It wouldn't be until 1999 that they released their first non-Amiga Toaster hardware if I'm not mistaken.
-
Yes, they really did bugger it up. When the A4000 and VT4000 (and then the Flyer) came out, production houses were clamoring to buy A4000's. I remember my local Amiga dealer griping about how much trouble he was having getting A4000 and A4000T's in, he was selling them quicker than he could get them. Compared to the competition, the VT was just so much of a better, and cheaper - solution.
All is not lost, though - Newtek is making money hand over fist with their TriCasters (among other products they offer), I suppose.
If you can credit one thing to Commodore, it is this - at every opportunity, it seems like they were compelled to cook their own Golden Goose rather than doing the smart thing and just collect the eggs.
-
Yeah, I'd like to give Commodore the benefit of the doubt as an engineering-centric company, which means they're predisposed to screwing themselves eventually if they don't get a reality check every once in a while. They get myopic and quite often don't really understand (or even value) the true end user experience with their creation or its weaknesses. That's like a cliche situation by now in the tech business.
I said in another thread, the situation with Toasters in 1994 and 1995 after all the Amiga 2000s and 4000s that were ever going to be produced had been produced was Toaster resellers were paying top dollar for dead A4000s (and dying in the first year was quite common, sadly, it happened to mine). One in LA was offering up to $1500 for an A4000 corpse, so that they could Frankenstein a few of these together into a working VT4000 workstation.
It was crazy. The demand was that high and there was just nothing else out there that was comparable for switching and DVE and transitions. I just never understood why this capability was suddenly so desirable now that it was more affordable than it had ever been. It was like the device created this need in people to figure out how they were going to use it rather than it being a solution to an outstanding need that wasn't being filled. It wasn't like all the other hardware you needed to actually make good video was suddenly cheaper.
The Toaster itself wasn't a panacea for prosumer video which needed to be TBC'd and was pushing a different spec for I/O. Hi-8mm, ED and Super-Beta and S-Video were ideally S-Video devices but the Toaster was composite only. Meanwhile, above the industrial threshold professional production was shifting to component for folks still working in analog and serial digital at the high end. The Toaster was therefore best matched to U-Matic (3/4") which was still a very ubiquitous format in actual TV production, college RTV departments and industrial production but well outside the budget of your average wedding shooter, with a single deck costing as much as an A4000 for a cheap, play-only model. NewTek reps were very coy and deflective when you started talking to them about whether or not they were going to offer additional models offering S-Video or component I/O. They'd supposedly designed it to be "D2 internal" which was an odd choice but no way of going in or out but composite analog.
By mid to late 1990s it got so you could recognize when a Toaster was being used on a program or video, whether or not they used the "Kiki Wipes" or those kitschy transitions with falling sheep and wiper blades, etc. What's sad is all this time, I'm looking at contemporary news programs in HD where, today, a lot of the character generator and lower-thirds and broadcast graphics being done aren't as good as what all the little cottage publishers were making to sell to Toaster users back then.
-
Wow, this turned out to be a book. Hopefully it even posts correctly.
A few years ago I asked Tim why there wasn't a PAL VT and he gave me a long story that I'll do my best to paraphrase.
The card shouldn't have worked in the first place. It was designed based on observation of what the components actually did in the real world, not the published specs. I'm not sure I have permission to go into details, but if you look at the spec sheets for the components, the Toaster isn't possible and a simple overclock won't fix it.
Also, back then the tools to make gate arrays were absolutely terrible, nothing like what we have now and the gate arrays were one of the keys to the whole thing working at all.
There weren't affordable tools to simulate the timing with the precision needed to handle the video signals and there weren't high-level languages to program them with anything non-trivial. Tim spent years hand tweaking every gate by hand to get the timing right and it was only right for NTSC. PAL would effectively be a complete rewrite and he never wanted to go through that ordeal again.
Personally I also think that he had proven it was possible, so the drive needed to expend that much effort just wasn't there anymore. From what I've seen, he's not a money guy, he's genuinely motivated by challenges and curiosity. Doing it again for PAL wouldn't be interesting at all.
As for "why Amiga?", other than everyone at NewTek being Amiga fanatics, it really was the only tool for the job. Years later, others were doing a subset of what the Toaster could do, but not the whole package in realtime and not with anything affordable. The Amiga chipset was incredibly flexible and more importantly, genlockable. The Toaster was designed around the Amiga to take advantage of it's strengths and even quirks. I was literally inseparable.
Truevision, etc had to essentially make their own "custom chipset" and add that to their cards too, driving up the R&D, testing and final shipping costs. The Amiga chipset was already stable, well documented and essentially free with the host computer. We could focus on what we did best. Without the chipset pushing animation data to the Toaster, who knows when a comparable card would ever be available.
We stuck with Amigas for years for much the same reason. The alternatives weren't up to the task. The other was that we not only loved the platform, but we already had a huge investment in software and hardware for it.
Somewhere just before the Flyer was about to ship, it was clear Commodore as we knew it was screwed. We even tried buying Commodore ourselves, but they were still priced crazily out of our reach.
At that point we started testing alternative platforms and even designed a sort of Draco-esque Amiga subset computer with our own custom Amiga-like OS. They had real CIAs, but other than that, all in house and off the shelf parts. If you've seen the tethered "PC Toaster" boxes with the video screen on the front panel that we showed at a few trade shows around 1995 or so, well, that's what's inside the box, a tiny passive backplane, a Flyer card in a Zorro slot and our own 68k(can't remember the model) computer on a card. The problem was that it wasn't better than the 4000 versions, so we scrapped it. I think I have the only surviving systems in my office, one of which still works.
Parallel to that we were writing theoretical code, counting clocks and running the numbers based on Moore's Law to see when it would be possible to do it on software. The hardware side of that project went on to become FrameFactory in about 1998, a basic uncompressed IO board that could bring in some much needed income since Amigas were few and far between.
It took a few more years for PCs to become fast enough, but a couple of years after that you could see us doing demos of what would eventually become VTNT. In the demos we were using dual 500MHz systems (the fastest we could get) and doing some pretty high latency examples of Toaster-like video.
It was almost 2002 before computer speed was adequate and the software was finally ready. When we released VT2, we finally a viable replacement.
Could we have designed a custom chip to do it earlier? Maybe, but at some point you have to pick a path and follow it to the end. Those years hurt, but the basics of the Frame Factory card and VT2 software carried on through the entire standard definition TriCaster line, which is an incredible lifespan by any measure and what we learned there is very much a part of what makes the current high-def TriCasterXD and 3Play lines possible.
-
A lot of this I don't buy. Someone who wasn't "a money guy" wouldn't have been so nouveau riche rap star "baller" about showing off Italian sports cars and jet planes. They had enough money to hire engineers to do the conversion and, if necessary, re-brute force the development of a PAL compatible version. They didn't. The rest seems like excuses. By his own description, his development process was more about being persistent than clever. That's actually the sort of thing that's easier for someone else to replicate, quicker.
The only thing the "chipset" actually provided was the genlockable character generation and scrolling text. It didn't do the DVE effects to the video itself. Amiga titling was generally better than the options available from PC, Mac and IIGS solutions. Of course that's not insignificant but even at double the price for a non-Amiga solution you'd still be looking at a significantly lower cost than just a professional, standalone character generator and those folks were used to paying more for their toys in any case.
A Truevision AT/Vista board was $3795 and that was mostly a paintbox and simple display buffer solution without the ability to process live video or playback the kind of animation that you got with the Toaster4000. It was a much better paintbox for sure than anything on the Amiga but that's just to illustrate the kind of pricing NewTek was competing with for much narrower focused devices. Double their Amiga prices would have still been a bargain. There wasn't anything "impossible" going on, as this was always religious hyperbole and folklore. The real question in the context of non-Amiga solutions was whether or not it could be done at the same price as an Amiga solution.
The info later in the post, talking about waiting for processors to get faster, etc. is obviously related more to being able to process the DVE effects on the host computer, take the video in, transform it, animate it, and then push it out. The Amiga VT didn't do that, there was an onboard DVE processor that did this, crudely in the first version and with a bit more polish in the VT4000 (ie. antialiasing). Having a software based DVE where the board itself is just handling the high speed video i/o from multiple sources is an entirely different beast than what the VT was and there's never been an Amiga that could have handled that sort of business with even standard definition video, I'm sorry.
The real story is they were loyal to the Amiga, beyond it's death until they had no alternative for continued survival.
-
@Heiroglyph
Quite a little "after market" thing here too:
http://www.tvworldwide.net/AboutUsYou/DaveGardy.aspx
Found this resume which also shows Dave:
http://www.geocities.ws/marypmiller/resume.htm
HUGE room full of Amigas running Lightwave, later still used for modelling & exporting scene files.
#6
-
Commodore really screwed the pooch here. This was, finally, their "killer app", one that compels people, often irrationally, to buy it and buy whatever computer is needed to run it. Like so many other "killer apps" you have a lot of purchases being made by people who have no real use for it, who won't really use it much at all much less to its fullest, but this is ultimately the goal and a practically necessary component for the economic health of media-based developers.
Problem is, this 'killer app' was 1) too expnsive for casual users and 2) too niche to effect sales enough.
As good as Video Toaster was, it was still a product for more serious Video professionals and smaller TV stations. This limited its customer base. If you look at other killer apps, like Visicalc, that was a piece of software that was within the budget of a lot more people and usefull to a wider range of customers.
There is no way Commodore could survive on Toaster sales alone, evidenced by them not. You also had Prevue channel using Amigas in cable systems to provide the channel guide and that didn't keep them from folding. Toaster, Lightwave and Prevue were there to the end of Commodore. Prevue being a big reason certain used Amigas were still well priced after Commodore went under since they were buying up many of them and Toaster being a big reason box Amigas remined valuable.
Commodore did not need a $4000 ($2,400 for the toaster, $1,500 for the 2000) killer app. I venture Newtek was probably making more money on every Toaster/Amiga setup that was sold than Commodore was. They needed a few hundred dollar killer app and to actually let people know about it and why they needed it and why they need an Amiga to run it for them. How it would help their daily lives, improve something they do all the time. That is what killer apps do.
-
A lot of this I don't buy. Someone who wasn't "a money guy" wouldn't have been so nouveau riche rap star "baller" about showing off Italian sports cars and jet planes. They had enough money to hire engineers to do the conversion and, if necessary, re-brute force the development of a PAL compatible version. They didn't. The rest seems like excuses. By his own description, his development process was more about being persistent than clever. That's actually the sort of thing that's easier for someone else to replicate, quicker.
All I've got to go on is the word of the guy who did it and the guy who built most of the Flyer. They have no reason to sugar coat it to a coworker and I believe them.
As for the money thing, who wouldn't buy some toys when you're a young guy with an overnight success beyond your wildest dreams?
The only thing the "chipset" actually provided was the genlockable character generation and scrolling text. It didn't do the DVE effects to the video itself. Amiga titling was generally better than the options available from PC, Mac and IIGS solutions. Of course that's not insignificant but even at double the price for a non-Amiga solution you'd still be looking at a significantly lower cost than just a professional, standalone character generator and those folks were used to paying more for their toys in any case.
That's not actually correct. Watch the control monitor of Toaster system when doing a transition. The animated effects are done by the Amiga, you just can't see them well because they are encoded with extra data that doesn't look correct on that monitor, somewhat similar to a DCTV image. It's a combination of overlay, matting and UV coordinate information that the Toaster card knows how to interpret. The gate arrays on the Toaster card then use that data to mix the video signals and potentially the decoded Amiga graphics.
Some slow control data (and grabbed images/CG sent to the Toaster) is sent through the parallel port on the video slot, but all the fast data is sent through the encoded digital video signal.
That's why the control screen "blanks" during certain operations.
A Truevision AT/Vista board was $3795 and that was mostly a paintbox and simple display buffer solution without the ability to process live video or playback the kind of animation that you got with the Toaster4000. It was a much better paintbox for sure than anything on the Amiga but that's just to illustrate the kind of pricing NewTek was competing with for much narrower focused devices. Double their Amiga prices would have still been a bargain. There wasn't anything "impossible" going on, as this was always religious hyperbole and folklore. The real question in the context of non-Amiga solutions was whether or not it could be done at the same price as an Amiga solution.
There's impossible and there is impossible with certain constraints, such as price and more personal constraints as "is this better".
We had some workable compromises, but if Tim didn't like it, it wasn't going out the door. It sucks sometimes, but NewTek has that sort of mentality for better or for worse. Personally I enjoy it. Most of the time ;)
The info later in the post, talking about waiting for processors to get faster, etc. is obviously related more to being able to process the DVE effects on the host computer, take the video in, transform it, animate it, and then push it out. The Amiga VT didn't do that, there was an onboard DVE processor that did this, crudely in the first version and with a bit more polish in the VT4000 (ie. antialiasing). Having a software based DVE where the board itself is just handling the high speed video i/o from multiple sources is an entirely different beast than what the VT was and there's never been an Amiga that could have handled that sort of business with even standard definition video, I'm sorry.
I've already addressed part of this above.
The Amiga Toaster had chips that processed video, but the mix was driven by the Amiga. The video didn't need to get to the host computer unless you grabbed a still, which is a very slow process on a Toaster.
Because the Amiga was naturally completely and reliably in sync with the video, this was an option.
When you try to do that in software, you have to get every frame all the way to the CPU and back to the card so a fast that a dumb card is potentially preferable. It's definitely a whole different beast because we couldn't think about the process the same way anymore. None of our old tricks were viable.
Going with smarter, more specialized hardware would also limit the lifespan and flexibility of the design. Those cards only became outdated when HD video became the norm.
The real story is they were loyal to the Amiga, beyond it's death until they had no alternative for continued survival.
No argument there.
-
Problem is, this 'killer app' was 1) too expnsive for casual users and 2) too niche to effect sales enough.
Photoshop was not an inexpensive app. Neither were any of the page layout tools that kept Macs relevant, none of which directly brought any dollars into Apple either. All of these very niche as well, yet they led to not only more people buying software they didn't actually need (this is the lifeblood of the industry overall) and machine purchases that allowed for the "one day" upward compatibility. The killer app is the tide that raises all boats. Because it's there, the hype surrounding it, all sorts of other smaller developers benefit. The killer app creates another tier of supporting apps. Commodore benefits from all that activity if they seize on it and support it and they didn't.
There was no existence in desktop publishing without spending thousands, in addition to whatever host computer you were using, and generally none of this additional expenditure went into the pocket of whomever made this computer.
Killer apps aren't about practical accessibility, or at least not exclusively so. They're about selling an idea and creating the feeling that you "needed" it. Since the dawn of the personal computer age most of this is irrational consumerism. The Toater is hardly the most expensive item people have lined up around the block to buy for their computers that they don't, ultimately, need.
-
A
That's not actually correct. Watch the control monitor of Toaster system when doing a transition. The animated effects are done by the Amiga, you just can't see them well because they are encoded with extra data that doesn't look correct on that monitor, somewhat similar to a DCTV image. It's a combination of overlay, matting and UV coordinate information that the Toaster card knows how to interpret. The gate arrays on the Toaster card then use that data to mix the video signals and potentially the decoded Amiga graphics.
Okay, but that's very sparse data. It's not heavy lifting at all. It's very clever, allowing new effects to be added without the need of firmware updates to the board. I would put money on this actually being less work than a full page of Kara Fonts characters scrolled from the bottom of the screen to the top. Some polygons and a few "magic bits" is hardly "Only Amiga Makes It Possible" material.
The Amiga Toaster had chips that processed video, but the mix was driven by the Amiga. The video didn't need to get to the host computer unless you grabbed a still, which is a very slow process on a Toaster.
You're using "mix" here as if it was actual work or some kind of special engineering. The video was processed on the board and sparse data sent from the Amiga instructed the board. It's not magic. If I'm watching Netflix streaming over a Chromecast device yet I'm controlling the Chromecast with an app on my phone I don't attribute any of the streaming or display or decompression, etc. to that mobile device in my hand.
That's a very simple analogy, simpler in many ways compared to what is happening in the Amiga to Toaster connection but there is nothing stopping an extension of this very sort of one-way relationship of control such that full scrubbing and editing and higher level application control is established in such a paradigm and no matter what I would never attribute any of the real work to the handheld device that's ultimately controlling and "driving the mix".
Going with smarter, more specialized hardware would also limit the lifespan and flexibility of the design. Those cards only became outdated when HD video became the norm.
Mmmmm, in local TV circles and the like, perhaps, which stay with hardware long passed its sell-by date. The Toaster became anachronistic with just a notch above the the more common industrial level (limping along on U-Matic) not long after the release of the VT4000. Because composite video. It would be the late 1990s but well before HD when the VT and its composite video became a real problem at the prosumer level. Because IEEE 1394, otherwise known as Firewire.
As soon as video acquisition and recording went digital the Toaster was pretty much public access television material. But by 1993 even, component video devices were making their way into prosumer hands even, via devices like the Personal Animation Recorder and later Perception Video Recorder. These rendered the framebuffer and video output on the Toaster a null and void issue for pretty much any application not involving the switching or processing of live video.
-
Okay, but that's very sparse data. It's not heavy lifting at all. It's very clever, allowing new effects to be added without the need of firmware updates to the board. I would put money on this actually being less work than a full page of Kara Fonts characters scrolled from the bottom of the screen to the top. Some polygons and a few "magic bits" is hardly "Only Amiga Makes It Possible" material.
I bet the efficient near real time multitasking of the AmigaOS had something to do with the "Only Amiga Makes It Possible". Wasn't the video output at normal speed in real time? Even fast Windows computers today will just lag for several seconds (or more) sometimes. This would kill anything real time. Even low end Amigas are more responsive than Windows PCs that are 1000 times faster. It's really sad that M$ can make an OS slower faster than new hardware makes it faster but then it allows them to sell more copies of their OS.
It's too bad C= didn't take advantage of the Toaster sales to bring down the cost of high end 68k processors with huge memory capacity and more expensive tower cases with more expandability. One question for Heiroglyph though. Why didn't LightWave ever get recompiled for the 68040 and 68060 (specifically the FPU)? This would have doubled the speed over using OxyPatcher. For being "Amiga lovers", they seemed happy enough to let LightWave processing move to the PC from a very early time.
-
Heiroglyph,
As a NewTek fan and user of a Toaster/Flyer setup (just something to play with these days) and SpeedEdit (more people should know about this NLE!), I just wanted to chime in and say I am enjoying your posts. Keep the stories and insights coming.
-
Tim Berners-Lee put up the first website on 06 August 1991. The World Wide Web is what drove popular video editing's popularity through the roof. Unfortunately it was too late for Amiga and too late for analogue video editors.
-
I bet the efficient near real time multitasking of the AmigaOS had something to do with the "Only Amiga Makes It Possible". Wasn't the video output at normal speed in real time?
I wish I could say yes, but that was something that really upset Commodore. We would often bypass the OS or prevent multitasking in order to get reliable throughput even on slower CPUs.
It was a win for users of 030's for example, but yeah, we weren't exactly system friendly.
We did love the way the OS worked though and when we moved on we modeled a lot of our internal code similarly.
Most people don't realize it, but the Flyer card is a self contained 68k computer that talks to the Amiga through the ZorroII bus. It runs its own Exec clone, much of which was also used in development of the small box I mentioned earlier. (yes, that little box has two independent 68k CPUs) You can find a version of our Exec and references to the stand alone box software in the Flyer part of the OpenVT source distribution.
One question for Heiroglyph though. Why didn't LightWave ever get recompiled for the 68040 and 68060 (specifically the FPU)? This would have doubled the speed over using OxyPatcher. For being "Amiga lovers", they seemed happy enough to let LightWave processing move to the PC from a very early time.
I honestly don't know. The 3D and video teams are very separate and I'm on the video side.
-
I wish I could say yes, but that was something that really upset Commodore. We would often bypass the OS or prevent multitasking in order to get reliable throughput even on slower CPUs.
It was a win for users of 030's for example, but yeah, we weren't exactly system friendly.
We did love the way the OS worked though and when we moved on we modeled a lot of our internal code similarly.
Most people don't realize it, but the Flyer card is a self contained 68k computer that talks to the Amiga through the ZorroII bus. It runs its own Exec clone, much of which was also used in development of the small box I mentioned earlier. (yes, that little box has two independent 68k CPUs) You can find a version of our Exec and references to the stand alone box software in the Flyer part of the OpenVT source distribution.
It's pretty amazing what the Amiga and NewTek could do with a 68030. Turning off multitasking during high CPU loads is understandable.
I honestly don't know. The 3D and video teams are very separate and I'm on the video side.
I disassembled LightWave 5.20a and I'm not sure they knew either. The optimization level is poor even for a 6888x. It's awful for a 68040 and 68060. I see instructions like this:
fdiv.w #2,fpn ; 6 bytes
which can be replaced exactly by:
fmul.s #0.5,fpn ; 8 bytes
The 6888x drops from ~130 cycles to ~90 cycles. The 68060 drops from 40 cycles to 4 cycles (the 68040 FPU is similar to the 68060). The code is littered with other F.w #imm,FPn instructions also which are slower to unnecessarily convert from int->fp. This does save a little code at the cost of speed but SAS/C also could not compress the double precision to single precision fp which would likely save more space overall. SAS/C is basically doing no scheduling of instructions to take advantage of the parallel operations of the integer and FPU units on all 68k FPUs either. Maybe the compilers weren't as good back then but there were still ways to do deal with these problems. My disassembly is clean enough I could spend a week optimizing and then reassemble with vasm's optimizer (vbcc's assembler) and probably double the speed on a 68040/68060 compared to OxyPatcher. More would be possible with an actual vbcc compile and my new fp math support. It would be interesting to see how much faster it would be with no traps at all. LightWave probably could have been 2-4 times faster on the 68040/68060 back then if NewTek just would have hired me. Ok, I wasn't that good fresh out of high school back then but it's too bad we can't compile the LightWave 5.20a sources with vbcc to get an idea of what the 68040 and 68060 were capable of back then (vbcc still doesn't have an instructions scheduler for the 68060 to really maximize performance).
-
I <3 all the history in this thread. Wish we could make it a sticky. :D