I realize I haven't posted in a while. Not even during the pre-release frenzy surrounding Frost, Krakatoa's BFF. Having spent the last 4 days in Las Vegas showing off Deadline, Krakatoa and Frost to anyone who came to see them, I thought it was time to write something about the future.
I am sure those of you who remember my "How Fast Is Fast" blog about Krakatoa 1.6 wouldn't expect yet another similar jump in performance from the next version. Well, you would be very very wrong, and I have the numbers to prove it. In fact, the upcoming version of Krakatoa might provide a bigger speed-up over 1.6 than 1.6 had over 1.5!
Creating Partitions and loading PRT file sequences from disk are among the most typical workflows in the Krakatoa pipeline, and the time to load the particles has traditionally been about half the time of the rendering process. Several factors affect the speed of loading - the speed of the hard drive or network connection serving the files, the speed of the CPU reading the data and decompressing the stream into memory, and the number of additional operations performed on the particles while loading, like MagmaFlow and Material evaluation, deformations and culling. In the past, the latter operations were gradually updated to support multi-threading, but the initial loading has remained limited to two threads, and typically saturated only one core.
Not anymore. I had the pleasure to benchmark an early alpha build of what might become Krakatoa 2.0 on a variety of machines with a multitude of storage solutions. To test the pure loading speed, I created a Box with dimensions 100x100x100, converted to a PRT Volume and partitioned it to disk as 100 partitions, each one with 1 million particles. Note that this introduces some overhead to the loading process - loading 10 partitions with 10 million each, or one partition with all 100 million would be somewhat faster, but I wanted to produce a more realistic case which is nearly the worst case of partitioning since Krakatoa currently limits the max. number of partitions to 100. Also, it would be able to create enough threads for any number of cores. I kept the default channels layout - Position, Velocity, Normals, Color, Density and ID - to simulate a typical case even though I did not need some of these like Velocity or ID for the actual rendering. I then created a single PRT Loader from all 100 partitions, created a default Spotlight and rendered.
My first test used the slightly outdated hardware tasked to perform the Thinkbox demos at the NAB show in Las Vegas - it was a dual Intel Core Duo, in other words four physical cores, no Hyperthreading. The interesting thing about this machine though was that it contained one 7200 RPM harddrive, two striped 10000 RPM drives, one SSD drive and a Fusion-io card, all connected to the same hardware. This gave me the ability to find out how the storage medium affects the new software.
For comparison, I used the current 1.6.1 build. Loading the 100 million particles with it took 57.3 seconds and the total rendering time was 2 minutes 46 seconds. It did not matter what drive I loaded the particles from because the speed was fully determined by the performance of the one core reading the ZIP stream from the PRT file.
Loading the particles from the 7200 RPM drive using the new build cut the loading time down to 38.2 seconds and the total rendering time to 2 minutes 29 seconds. This is not a very impressive speed-up, but it reached the physical limitations of the hard drive, while loading the CPUs as much as the I/O bottleneck allowed. Having 4 cores, 4 threads were created to read 4 PRT streams at once, but the drive could not keep up with the demand.
Loading the particles from the two 10000 RPM drives brought the 4 CPUs to about 80% saturation before the bandwidth of the hard disks became the bottleneck again. The time to load the 100MP went down to impressive 16.1 seconds, but it was obvious that there was more to be expected from the solid state drives. And indeed, running the exactly same tests from the SSD drive gave me 11.2 seconds for loading and 1 minute 49.9 seconds total render time, while saturating all 4 cores completely! Trying the same with the even faster Fusion-io card produced the same loading and rendering time, clearly proving I had reached the CPU bottleneck.
Thankfully, Fusion-io was well represented at NAB and I got the chance to run the benchmark on an 8 core machine to test the waters. My gut feeling told me I should expect about half the loading time with twice as many cores so I wasn't exactly surprised when the faster system loaded the 100 million particles in 6 seconds and finished rendering in only 52 seconds! (Un)fortunately, all 8 cores were once again at 100%, making the result CPU-bound instead of I/O bound, leaving me wanting to test on a 16 or 32 core machine to see what a Funsion-io card can really do for Krakatoa. My gut feeling tells me again we could expect loading times of 3 seconds or less for 100 million particles on such a system, but until I actually get to try one, I can live with the results pretty well. Supposedly, a good SSD drive could keep up with an 8 core system to produce the 6 seconds loading time, too, so you don't have to spend the equivalent of a new car to get that performance...
There are several other areas that have seen some speed up in the new version - on my home i7 quadcore machine, the sorting for both lighting and drawing of 100 MP went down from 12 seconds to about 8, and the drawing was reduced from 5 to 4 seconds. That machine is not very good for testing the loading improvements due to a slow hard drive though, so if you could imagine a modern computer with a lot of cores and fast SSD drives, Krakatoa will literally fly on it later this year!
Obviously, this is just the tip of the iceberg when it comes to what will be new in the next version of Krakatoa. Wait for Siggraph and be very, very excited - I know I am...
Friday, April 15, 2011
Saturday, September 25, 2010
Servicing the Pack
The Krakatoa 1.6.0.43178 build released on August 31st is no more. It has been replaced by a 1.6.0.43376 build released on September 24th. If you are already running 1.6.0, you should do yourself a favor and update now. Go, I will be waiting!
This is the first time we replace a build without incrementing the version number. We call it a Service Pack 1, but you can read it as "The build we should have released to start with" :) It fixes several bugs related to rendering FumeFX, adds the Icon Size spinners to all PRT objects that were missing one in the original build, and because there have been more than three weeks since the release, it also adds some new features here and there - Yours Truly cannot sit around without touching up the user interface...
Here is the list of changes:
Bug Fixes
Feature Improvements
User Interface
Just the first 4 points of the Bug Fixes list should be a good reason to update if you are using FumeFX 2.0, Thinking Particles or modifiers on PRT Loaders. In fact, I had to postpone a video tutorial recording session I had planned because it totally depends on a modifier on a PRT Loader and would behave quite strangely in the original 1.6.0 release... Stay tuned! And if you haven't watched the other YouTube videos, go here, here, here and here.
If you find more bugs hiding in the newest build, please let us know. We will chase them and hit them with a big hammer!
This is the first time we replace a build without incrementing the version number. We call it a Service Pack 1, but you can read it as "The build we should have released to start with" :) It fixes several bugs related to rendering FumeFX, adds the Icon Size spinners to all PRT objects that were missing one in the original build, and because there have been more than three weeks since the release, it also adds some new features here and there - Yours Truly cannot sit around without touching up the user interface...
Here is the list of changes:
Bug Fixes
- PRT FumeFX with FumeFX 2.0 was causing a crash when switching to Post Cache mode. This has been fixed.
- PRT FumeFX was showing only a portion of the simulation due to incorrect handling of Adaptive bounds. This has been fixed.
- Thinking Particles was causing particle evaluation to hang at render time if there were two or more TP objects in the scene. Fixed.
- Deformation Modifiers were inverting the Velocity Channel in PRT Loaders. This has been fixed.
- Fixed the MAXScript call for saving all rollouts to presets which was causing a System Exception.
- Fixed a crash when trying to load a Save Channels Preset.
- Turning off the Viewp.Spacing in PRT Volume causes the render-time Spacing value to be used for the viewport, too, but changing the spinner wasn't updating the viewport cache and required a manual update. This has been fixed.
Feature Improvements
- The Thinking Particles Partitioning code was tweaked to allow an Integer Helper containing "*RandomSeed*" in its name to be incremented just like a regular "RandomSeed" property. This can be used to adjust the Random Seed of MatterWaves by connecting an Integer Helper node to the RandomSeed property which is otherwise not exposed to MAXScript.
- Undo/Redo wasn't restoring the Input Nodes' value correctly. This has been fixed. Please note that if the track is keyframed, performing Undo/Redo can affect the complete curve as the animation state at the time of Undo creation is not stored and there is no way to store/restore the complete controller. This fix also improves the Macro Recorder playback.
- The ">Interactive Mode (SLOW!)" option in the Krakatoa Channels Modifier wasn't handling Geometry objects used in Surface Operators. This has been fixed.
- Added conversion buttons to the SurfDataValue Operator to quickly turn the output value from Integer to Float, from Float to Vector, extract X,Y and Z components of Vector or transform Vector From World to Object space or into View Space.
- Added an explicit INSERT/BRANCH mode displayed in the title bar of the MagmaFlow Editor and controlled by the Insert key in the Numeric Keypad. When set to Branch and a single node is selected that has an output socket connection, creating a new operator will not insert into the existing connection but branch into a new output connection. Once the node is branched, the mode will revert to INSERT automatically, since it makes more sense most of the time. In previous versions, this was done with the SHIFT key, but it could collide with some keyboard shortcuts that use the SHIFT key as a modifier key.
- Fixed the mouse right-click hit testing when the MagmaFlow Editor is docked as Extended Viewport.
- Added support for Schematic Flow as Extended Viewport, including correct right-click handling.
User Interface
- PRT Volume, PRT FumeFX and PRT Source now expose an Icon Size spinner to the UI. The MacroScripts now use the XY size to define the icon size of PRT Volume and PRT FumeFX.
- Added support for saving and loading the Channels To Save list.
- Added all rollouts to the right-click menu of the Float/Dock icons with a check mark for visibility. Selecting will toggle visibility, holding SHIFT will navigate to the rollout.
- Added right-click handling to all "Back To Main Controls..." buttons in addition to the left-clicking.
- Added the keyboard shortcut names to the right-click menu of the Float/Dock icons.
- Changed the navigation buttons throughout the UI to unhide hidden rollouts before navigating to them.
- Reworked all MacroScripts to disable themselves and avoid any crashes if the Krakatoa path is removed from the plugin.ini.
- The Rollout Preset files were not restoring the visibility of the "Shader Parameters" rollout correctly - even if saved as hidden, it would be displayed because the update of "Main Controls" rollout's Shading Mode list was forcing it to appear. This has been fixed by processing the "Main Controls" before "Shader Parameters".
- Optimized the "Load From History" dialog to use the History Cache instead of scanning the files from disk. With 8281 History files on disk, the old code was taking 47 seconds to update the first time the dialog was opened, and 10 seconds in consecutive attempts. With the new code, the time to open the dialog went down to 0.078 seconds!
Just the first 4 points of the Bug Fixes list should be a good reason to update if you are using FumeFX 2.0, Thinking Particles or modifiers on PRT Loaders. In fact, I had to postpone a video tutorial recording session I had planned because it totally depends on a modifier on a PRT Loader and would behave quite strangely in the original 1.6.0 release... Stay tuned! And if you haven't watched the other YouTube videos, go here, here, here and here.
If you find more bugs hiding in the newest build, please let us know. We will chase them and hit them with a big hammer!
Friday, September 24, 2010
I Am A Cheater, Too!
Thanks to this CG Society article, I discovered the following "Making Of Tron" video on YouTube:
Part 1 Part 2 Part 3 Part 4 Part 5 Part 6 Part 7 Part 8 Part 9
The video is beyond awesome, but the most striking part of it, even more than the behind the scenes footage and image material, was THIS SENTENCE by Steven Lisberger:
"The Motion Picture Academy refused to nominate Tron for special effects because they said we cheated when we used computers..."
While I point at Star Wars as the main influence for me to end up in the visual effects industry, Tron has a very special place in my heart, too. Unlike Star Wars, Tron was not shown in communist Bulgaria's movie theaters back in the early 80s. But the magazine for international journalism "Parallels" ("Paraleli" in Bulgarian) had a large article with images about the making of the "first computer generated film" which I used for a school project about "an art form of my choice" (film is considered the Seventh Art Form, so I went with Science Fiction Movies as my theme).
For 6 years, I could only dream about seeing the actual film. My dream came true thanks to "Perestroika" on February 14th 1988, a Sunday. It was 3 days before my birthday and it happened to be a weekend. I was in my first year of army service and my girlfriend came to visit me so I got a two days leave. We went to a so-called "Cinema Video Club" - these were coffee shops with video projector that were showing private (read: illegally copied) VHS records of western movies that could not be seen or obtained through any other channels. Copyright law did not extend beyond the Iron Curtain, and thankfully the idea that western movies were "capitalist propaganda designed to brainwash the youth" was not so popular anymore, so we could finally see films we had missed originally. And they were showing Tron that day!
Of course, I was completely blown away by what I saw then. A year later, during my second year of military service, I was in a computer center and had the opportunity to play with simple computer graphics. This led to getting a PC home with me for another two years after I finished service as I continued working as consultant to one department of the military academy. Then, in the first year as student of architecture, I presented a design project that was made and printed on a computer. Guess what the reaction of my professors was? "This is not fair to your peers, you cheated by using a computer".
As I have mentioned in previous blogs, I never intended to turn computer graphics into profession, mainly because it was impossible in the current political climate. But for some people who were living in a different environment, Tron turned out to be the actual inspiration for working in the visual effects industry.
There still is an original Tron coin-op game in one of our offices. This is because the company previously known as Frantic Films was started by Chris Bond who was one of those people who were inspired by Tron and actually went into the visual effects industry because of that inspiration. With Tron Legacy less than three months away, it is both a wonderful and curious feeling to look back at where things started and realize how far we, the industry and the world have come...
Part 1 Part 2 Part 3 Part 4 Part 5 Part 6 Part 7 Part 8 Part 9
The video is beyond awesome, but the most striking part of it, even more than the behind the scenes footage and image material, was THIS SENTENCE by Steven Lisberger:
"The Motion Picture Academy refused to nominate Tron for special effects because they said we cheated when we used computers..."
While I point at Star Wars as the main influence for me to end up in the visual effects industry, Tron has a very special place in my heart, too. Unlike Star Wars, Tron was not shown in communist Bulgaria's movie theaters back in the early 80s. But the magazine for international journalism "Parallels" ("Paraleli" in Bulgarian) had a large article with images about the making of the "first computer generated film" which I used for a school project about "an art form of my choice" (film is considered the Seventh Art Form, so I went with Science Fiction Movies as my theme).
For 6 years, I could only dream about seeing the actual film. My dream came true thanks to "Perestroika" on February 14th 1988, a Sunday. It was 3 days before my birthday and it happened to be a weekend. I was in my first year of army service and my girlfriend came to visit me so I got a two days leave. We went to a so-called "Cinema Video Club" - these were coffee shops with video projector that were showing private (read: illegally copied) VHS records of western movies that could not be seen or obtained through any other channels. Copyright law did not extend beyond the Iron Curtain, and thankfully the idea that western movies were "capitalist propaganda designed to brainwash the youth" was not so popular anymore, so we could finally see films we had missed originally. And they were showing Tron that day!
Of course, I was completely blown away by what I saw then. A year later, during my second year of military service, I was in a computer center and had the opportunity to play with simple computer graphics. This led to getting a PC home with me for another two years after I finished service as I continued working as consultant to one department of the military academy. Then, in the first year as student of architecture, I presented a design project that was made and printed on a computer. Guess what the reaction of my professors was? "This is not fair to your peers, you cheated by using a computer".
As I have mentioned in previous blogs, I never intended to turn computer graphics into profession, mainly because it was impossible in the current political climate. But for some people who were living in a different environment, Tron turned out to be the actual inspiration for working in the visual effects industry.
There still is an original Tron coin-op game in one of our offices. This is because the company previously known as Frantic Films was started by Chris Bond who was one of those people who were inspired by Tron and actually went into the visual effects industry because of that inspiration. With Tron Legacy less than three months away, it is both a wonderful and curious feeling to look back at where things started and realize how far we, the industry and the world have come...
Wednesday, September 1, 2010
Shipping Time
Krakatoa v1.6.0 is here! Where? There!
Surprisingly, we actually hit the shipping deadline this time. Not that it would have mattered one way or another since all users with support contract and even anyone who asked for it had access to the Beta builds for months. In the past, we always went by the saying "It will ship when it is ready". Being a niche product developed by a small team, we always had that luxury large publicly traded companies do not have - we'd rather delay a release than ship with known bugs. That's not to say there are no unknown bugs in there - if you find any, please tell us!
As mentioned in previous blogs, Krakatoa is now heavily multi-threaded and thus a whole lot faster. It integrates better with FumeFX incl. v2.0, supports the shipping version of Thinking Particles 4, provides new MagmaFlow operators, produces some Render Elements, generates much better shadows and fixes a lot of shortcoming of version 1.5.1. The UI is faster and more streamlined, while the number of additional tools has gone up. Look out for the Krakatoa menu in the Main Menu bar of 3ds Max - it exposes most tools that required customized toolbar icons in the past.
The new version even comes with a new license dialog written specifically to replace the totally unusable FlexLM one that has plagued humanity for years. Yes, the number one support problem of Krakatoa was caused by a piece of code that wasn't written by us. We really hope entering the license server or file will be much easier now...
On the curious side, possibly the smallest change with the largest impact on usability (or at least my sanity as a user) was the addition of a separator between the Copy and Clear menu items of the Log Window's Edit menu. You have no idea how many times I have cleared the window instead of copying the selection in previous versions!
We committed some "crimes", too. A day before release, a user asked for better PRT object icons to easier distinguish between PRT Loaders, PRT Volumes and PRT FumeFX objects. The request came literally in the last minute and no developers in their right mind would start changing that amount of code between Release Candidate 2 and Release Build. Well, fearless team that we are, we designed new icons, wrote a script to export their meshes into C++ compatible vertex and face lists and submitted the new code AN HOUR before the final build was produced! The impact risk was low, the positive effect on usability was significant. We had discussed that change before, but there were so many more important features and fixes, we never found the time. When everything else was done, it felt like we could afford to sneak that in and make some people happy...
Now that we have the final build out there, I really hope to be able to produce some new videos - my first and last Krakatoa-related videos were recorded using v1.0.0 back in 2007!
You can read more about the features of Krakatoa v1.6.0 here and here.
Surprisingly, we actually hit the shipping deadline this time. Not that it would have mattered one way or another since all users with support contract and even anyone who asked for it had access to the Beta builds for months. In the past, we always went by the saying "It will ship when it is ready". Being a niche product developed by a small team, we always had that luxury large publicly traded companies do not have - we'd rather delay a release than ship with known bugs. That's not to say there are no unknown bugs in there - if you find any, please tell us!
As mentioned in previous blogs, Krakatoa is now heavily multi-threaded and thus a whole lot faster. It integrates better with FumeFX incl. v2.0, supports the shipping version of Thinking Particles 4, provides new MagmaFlow operators, produces some Render Elements, generates much better shadows and fixes a lot of shortcoming of version 1.5.1. The UI is faster and more streamlined, while the number of additional tools has gone up. Look out for the Krakatoa menu in the Main Menu bar of 3ds Max - it exposes most tools that required customized toolbar icons in the past.
The new version even comes with a new license dialog written specifically to replace the totally unusable FlexLM one that has plagued humanity for years. Yes, the number one support problem of Krakatoa was caused by a piece of code that wasn't written by us. We really hope entering the license server or file will be much easier now...
On the curious side, possibly the smallest change with the largest impact on usability (or at least my sanity as a user) was the addition of a separator between the Copy and Clear menu items of the Log Window's Edit menu. You have no idea how many times I have cleared the window instead of copying the selection in previous versions!
We committed some "crimes", too. A day before release, a user asked for better PRT object icons to easier distinguish between PRT Loaders, PRT Volumes and PRT FumeFX objects. The request came literally in the last minute and no developers in their right mind would start changing that amount of code between Release Candidate 2 and Release Build. Well, fearless team that we are, we designed new icons, wrote a script to export their meshes into C++ compatible vertex and face lists and submitted the new code AN HOUR before the final build was produced! The impact risk was low, the positive effect on usability was significant. We had discussed that change before, but there were so many more important features and fixes, we never found the time. When everything else was done, it felt like we could afford to sneak that in and make some people happy...
Now that we have the final build out there, I really hope to be able to produce some new videos - my first and last Krakatoa-related videos were recorded using v1.0.0 back in 2007!
You can read more about the features of Krakatoa v1.6.0 here and here.
Tuesday, July 20, 2010
Changes and Dependencies
A couple of days ago somebody sent me a link to an automatic "Siggraph Paper Title Generator" which produced quite convincing scientific-sounding output. The title of this post could easily be mistaken for something programming-related, but it is about something else completely.
I haven't posted since April because the last couple of months were quite crazy. In short, my family just finished relocating through half Canada - from Winnipeg, Manitoba (the city closest to the geometric center of North America) to Vancouver, British Columbia (at the very edge of the continent). This is a quite positive change - no more winter temperatures lower than on the South Pole, no more mosquitos in the summer (considered the "national bird" in Manitoba ;o)) And last but not least, finally some hills, mountains and curves on the roads to make the driver in me happy.
The move itself went ok without any incidents, all our stuff arrived a week later than expected but absolutely preserved (especially the PC and the TV - sigh of relief), and the first thing after ensuring electric power was still on was calling the cable/internet provider to order the Most Important Thing after water and bread. This was on June 28th and I was told that the earliest possible appointment date would be July 11th. TWO WEEKS without Web, Email and Skype? Oh my...
The very next day I replaced my trusty dumb phone with a smart one just to get some form of online access. This gave us at least some email, web services and the ability to chat via text messaging with our parents in Europe, plus the added bonus of GPS turn-by-turn navigation. While this was one of the best purchases considering the situation, it wasn't enough and I finally realized how dependent (read: Addicted!) we have become to the internet.
Two days after arrival I was chatting with a colleague and discovered that another company otherwise known for its wireless services and cute animal commercials had started to provide optical cable in our area. In fact, they even had a brochure sent to the "current occupant" of our apartment, but my selective reading brain totally skipped over the first two letters of the title (which said "TV"!) that also promised phone and internet. So I immediately called the company and asked if we can even get their services, since I knew that some large buildings downtown had preferred partnership with the former cable TV monopolist I already had an appointment with. The answer I received left me nearly speechless. "Looking at our data here, it appears that your building has been pre-wired for out optical cable services, and TV and internet are already pre-paid by the building management for a year!". Oh, and they could come 5 days earlier than the competition (YAY!)
Too good to be true?
The lady on the phone promised to call me half an hour later to confirm my credit check and the appointment for the 6th of July. She never did, so the next day I called again. The person at the other end of the line went through the whole identification process and confirmed the 6th of July appointment. So on the 6th of July I took an hour off from work and went home to be around during the installation... except that nobody showed up. Another phone call later I learned the amazing truth: Computers Hate Me!
I was told that a software bug in the internal system of the company is preventing anyone in the call center from scheduling that appointment (three people so far had attempted to enter it and all got the error which I was told "nobody had seen before"). Being a TD and fighting the good fight with bugs every day, I don't really mind bugs, but I do mind when people don't tell me about them (Funnily enough, the name of the company implies they would "tell us" when something like that happens ;o)). So I was told that the top-dogs of the software team were looking into the problem and would call me as soon as it was fixed. I was even encouraged to call again the next day to check. Of course I did and had my fifth conversation with a fifth person about the whole issue. "Please believe us that we are working on fixing the bug". In all this time, I never received a call back from anyone. Probably because nobody cared personally, or probably because my bill was already paid for a year anyway, so an important stimulus was missing... ;o)
While getting home from work via a slightly different route that day, I realized that the head quarters of the company I was dealing with was literally the NEXT BUILDING to ours! So the next morning I went into the lobby and asked to talk to someone responsible for customer relations. Instead, I got connected with the technical team responsible for that "bug" and was told the bug was fixed and they could finally schedule an appointment. (Until this day I don't know whether this was just good timing or there was never a real bug). Of course, this was on July 9th so the earliest possible opening was for July 16th - over two weeks after my first call...
Being as addicted as I turned out to be, I could not wait 5 more days without internet. So we scheduled the 16th but also left the initial appointment with the other company on the 11th and on that glorious Sunday finally got some cable to scratch that online itch. Four days later we switched to the optical cable with HD, free PVR, 15 Mbps wireless access from all devices and finally felt home... This blog is just a reminder how much our lives have changed in the last decade and how dependent we have become... Now let's hope SkyNet won't take over soon ;o)
Back on topic (this is a Krakatoa/MAXScript blog after all) - the next shocker is that the year of the 20th birthday of 3D Studio will be the first in 11 years with me missing Siggraph. I wish everyone going a great time and while I am sad we won't be able to show you first hand what we have done to Krakatoa 1.6, I promise we will be very active online once it is released. I am spending a lot of time polishing both the documentation and the UI, and a lot of features are still being tweaked for maximum flexibility.
I hope the next blog will be a lot more interesting.
I haven't posted since April because the last couple of months were quite crazy. In short, my family just finished relocating through half Canada - from Winnipeg, Manitoba (the city closest to the geometric center of North America) to Vancouver, British Columbia (at the very edge of the continent). This is a quite positive change - no more winter temperatures lower than on the South Pole, no more mosquitos in the summer (considered the "national bird" in Manitoba ;o)) And last but not least, finally some hills, mountains and curves on the roads to make the driver in me happy.
The move itself went ok without any incidents, all our stuff arrived a week later than expected but absolutely preserved (especially the PC and the TV - sigh of relief), and the first thing after ensuring electric power was still on was calling the cable/internet provider to order the Most Important Thing after water and bread. This was on June 28th and I was told that the earliest possible appointment date would be July 11th. TWO WEEKS without Web, Email and Skype? Oh my...
The very next day I replaced my trusty dumb phone with a smart one just to get some form of online access. This gave us at least some email, web services and the ability to chat via text messaging with our parents in Europe, plus the added bonus of GPS turn-by-turn navigation. While this was one of the best purchases considering the situation, it wasn't enough and I finally realized how dependent (read: Addicted!) we have become to the internet.
Two days after arrival I was chatting with a colleague and discovered that another company otherwise known for its wireless services and cute animal commercials had started to provide optical cable in our area. In fact, they even had a brochure sent to the "current occupant" of our apartment, but my selective reading brain totally skipped over the first two letters of the title (which said "TV"!) that also promised phone and internet. So I immediately called the company and asked if we can even get their services, since I knew that some large buildings downtown had preferred partnership with the former cable TV monopolist I already had an appointment with. The answer I received left me nearly speechless. "Looking at our data here, it appears that your building has been pre-wired for out optical cable services, and TV and internet are already pre-paid by the building management for a year!". Oh, and they could come 5 days earlier than the competition (YAY!)
Too good to be true?
The lady on the phone promised to call me half an hour later to confirm my credit check and the appointment for the 6th of July. She never did, so the next day I called again. The person at the other end of the line went through the whole identification process and confirmed the 6th of July appointment. So on the 6th of July I took an hour off from work and went home to be around during the installation... except that nobody showed up. Another phone call later I learned the amazing truth: Computers Hate Me!
I was told that a software bug in the internal system of the company is preventing anyone in the call center from scheduling that appointment (three people so far had attempted to enter it and all got the error which I was told "nobody had seen before"). Being a TD and fighting the good fight with bugs every day, I don't really mind bugs, but I do mind when people don't tell me about them (Funnily enough, the name of the company implies they would "tell us" when something like that happens ;o)). So I was told that the top-dogs of the software team were looking into the problem and would call me as soon as it was fixed. I was even encouraged to call again the next day to check. Of course I did and had my fifth conversation with a fifth person about the whole issue. "Please believe us that we are working on fixing the bug". In all this time, I never received a call back from anyone. Probably because nobody cared personally, or probably because my bill was already paid for a year anyway, so an important stimulus was missing... ;o)
While getting home from work via a slightly different route that day, I realized that the head quarters of the company I was dealing with was literally the NEXT BUILDING to ours! So the next morning I went into the lobby and asked to talk to someone responsible for customer relations. Instead, I got connected with the technical team responsible for that "bug" and was told the bug was fixed and they could finally schedule an appointment. (Until this day I don't know whether this was just good timing or there was never a real bug). Of course, this was on July 9th so the earliest possible opening was for July 16th - over two weeks after my first call...
Being as addicted as I turned out to be, I could not wait 5 more days without internet. So we scheduled the 16th but also left the initial appointment with the other company on the 11th and on that glorious Sunday finally got some cable to scratch that online itch. Four days later we switched to the optical cable with HD, free PVR, 15 Mbps wireless access from all devices and finally felt home... This blog is just a reminder how much our lives have changed in the last decade and how dependent we have become... Now let's hope SkyNet won't take over soon ;o)
Back on topic (this is a Krakatoa/MAXScript blog after all) - the next shocker is that the year of the 20th birthday of 3D Studio will be the first in 11 years with me missing Siggraph. I wish everyone going a great time and while I am sad we won't be able to show you first hand what we have done to Krakatoa 1.6, I promise we will be very active online once it is released. I am spending a lot of time polishing both the documentation and the UI, and a lot of features are still being tweaked for maximum flexibility.
I hope the next blog will be a lot more interesting.
Sunday, April 18, 2010
How Fast is Fast?
In pretty much every review or comment about Krakatoa, I see something along the lines of "Krakatoa renders really fast". Don't get me wrong, I like reading that, but let's be fair - speed is relative (ask Albert E.) and people don't say this because Krakatoa is actually fast, but because most mesh-based particle rendering solutions for 3ds Max are slower in comparison.
All this good press makes my job harder as I try to push for speed improvements in the development of future Krakatoa versions (you might not know this, but my middle name is "Speeeed!"). We are in the business of selling software to users, and if a single copy of Krakatoa can get the job on time, why buy more? There was even an anecdotal case where an important customer of Krakatoa wanted to buy additional licenses, but before the transaction could be finalized, they told us they got their project finished with the few seats they already had because "Krakatoa renders so fast". Obviously, while a fast product is good for the image of the product, it could be quite detrimental to sales - thankfully, the client still bought those additional licenses so no harm done, but it made us thinking...
What would a company do if its existence depended solely on software sales? Things like limiting the Evaluation version to do a lot less (most of the particle management features of Krakatoa like partitioning, converting particle files between formats, deforming and modifying, importing back into Particle Flow etc. are basically free) or making Krakatoa deliberately slower come to mind. Thankfully, the main reason for Krakatoa's existence is the internal need of our visual effects production for a fast particle renderer. And watching Krakatoa in real-world production proves my point - Krakatoa can be damn slow and we have to do something about it!
So how fast is fast? Some people might answer "real time on the GPU would be good enough". Unfortunately, this is not the right answer. There have been some examples of real time CUDA-accelerated particle rendering in nVidia demos, but this is not where we are going and the reason is the data amounts we deal with. The main objective of Krakatoa is the "fast rendering of vast amounts of particles". In this objective, the second part has higher importance than the first one. With the arrival of 64 bit computing and the increases of installed RAM, our workstations can typically fit around 700 MP (Million Particles, or MegaPoints) in 16 GB of RAM, and we often go there or near. A typical graphics card with 1GB of RAM can handle less than 1/16th of that amount, so we don't care about that approach just yet, although it would be great for fast tests with a fraction of the particle count.
Another thing to keep in mind is that various areas of Krakatoa have different impact on speed - some are as fast as we could make them, some could be made faster, some can become bottlenecks depending on the settings. For example, the loading of particles which also includes the evaluation of materials, maps, deformation modifiers, MagmaFlows and particle culling was sped up in v1.5.0 via multi-threading and is now several times faster than in 1.1.x. The sorting portion has been multi-threaded for years and is probably as fast as it could be. The drawing in the Lighting and the Final Passes of Particle Rendering has remained single-threaded since the first version of Krakatoa and has been measured at around 2 MP/second. Given the increasing number of cores in today's machines, this is an area that could improve a lot! The evaluation and processing of Matte Objects was also sped up in 1.5.0 by simply switching from a raytracer to a rasterizer, but it could also benefit from multi-threading the depth map generation.
Adding a new light adds another sorting pass (already fast) and another drawing pass (not as fast as it could be). The moment you check the Motion Blur option, you ask Krakatoa to draw the Final Pass several times. If the number of samples is 8, Krakatoa has to do 8 times the drawing work and typically scales linearly (it actually takes 8 times longer to draw the particles if you ignore the loading time or render with PCache and LCache enabled). But if you have 8 cores in your workstation and Krakatoa would use them all to speed up the drawing, it would mean that 8 passes motion blur would "cost" as much as one pass right now. Wouldn't that be great?
You betcha! So that's what the next version of Krakatoa will do (and more). And the more particles you throw at it, the better it will scale. On my workstation, the pure drawing speed of 100 million particles comes close to 17.5 MP/second! The generation of a Matte depth map from 100 million polygons, something that is also performed before each motion blur pass, went down from 32 seconds to 4 (8x4=32, you know?). And we are not even half-done yet. Add to that the loading speed up with PRT Volumes already reported in a previous blog and you will see how this new build is shaping up as the fastest Krakatoa you have ever seen. In the same blog I mentioned it might ship as v1.6.0 - that suspicion turned out true and we already started updating the documentation for this upcoming 1.6.0 build which should be expected sometime before Siggraph.
I am quite excited about these improvements which we haven't even passed to the Beta testers yet as they happened in the last 10 days. All this wouldn't have been possible without our relentless production team which not only pushes Krakatoa to places it has never been to yet, but also constantly bitches about how slow it is! Love you guys! :)
So don't panic, we won't make Krakatoa slower just to sell more network render licenses and Deadline seats to run them, or disable more features so you cannot play with it for free. We understand that a faster Krakatoa is easier to love, and a Krakatoa you love is one you would if not buy yourself, then at least suggest to your employer (the guy with the wallet ;))...
Stay tuned!
All this good press makes my job harder as I try to push for speed improvements in the development of future Krakatoa versions (you might not know this, but my middle name is "Speeeed!"). We are in the business of selling software to users, and if a single copy of Krakatoa can get the job on time, why buy more? There was even an anecdotal case where an important customer of Krakatoa wanted to buy additional licenses, but before the transaction could be finalized, they told us they got their project finished with the few seats they already had because "Krakatoa renders so fast". Obviously, while a fast product is good for the image of the product, it could be quite detrimental to sales - thankfully, the client still bought those additional licenses so no harm done, but it made us thinking...
What would a company do if its existence depended solely on software sales? Things like limiting the Evaluation version to do a lot less (most of the particle management features of Krakatoa like partitioning, converting particle files between formats, deforming and modifying, importing back into Particle Flow etc. are basically free) or making Krakatoa deliberately slower come to mind. Thankfully, the main reason for Krakatoa's existence is the internal need of our visual effects production for a fast particle renderer. And watching Krakatoa in real-world production proves my point - Krakatoa can be damn slow and we have to do something about it!
So how fast is fast? Some people might answer "real time on the GPU would be good enough". Unfortunately, this is not the right answer. There have been some examples of real time CUDA-accelerated particle rendering in nVidia demos, but this is not where we are going and the reason is the data amounts we deal with. The main objective of Krakatoa is the "fast rendering of vast amounts of particles". In this objective, the second part has higher importance than the first one. With the arrival of 64 bit computing and the increases of installed RAM, our workstations can typically fit around 700 MP (Million Particles, or MegaPoints) in 16 GB of RAM, and we often go there or near. A typical graphics card with 1GB of RAM can handle less than 1/16th of that amount, so we don't care about that approach just yet, although it would be great for fast tests with a fraction of the particle count.
Another thing to keep in mind is that various areas of Krakatoa have different impact on speed - some are as fast as we could make them, some could be made faster, some can become bottlenecks depending on the settings. For example, the loading of particles which also includes the evaluation of materials, maps, deformation modifiers, MagmaFlows and particle culling was sped up in v1.5.0 via multi-threading and is now several times faster than in 1.1.x. The sorting portion has been multi-threaded for years and is probably as fast as it could be. The drawing in the Lighting and the Final Passes of Particle Rendering has remained single-threaded since the first version of Krakatoa and has been measured at around 2 MP/second. Given the increasing number of cores in today's machines, this is an area that could improve a lot! The evaluation and processing of Matte Objects was also sped up in 1.5.0 by simply switching from a raytracer to a rasterizer, but it could also benefit from multi-threading the depth map generation.
Adding a new light adds another sorting pass (already fast) and another drawing pass (not as fast as it could be). The moment you check the Motion Blur option, you ask Krakatoa to draw the Final Pass several times. If the number of samples is 8, Krakatoa has to do 8 times the drawing work and typically scales linearly (it actually takes 8 times longer to draw the particles if you ignore the loading time or render with PCache and LCache enabled). But if you have 8 cores in your workstation and Krakatoa would use them all to speed up the drawing, it would mean that 8 passes motion blur would "cost" as much as one pass right now. Wouldn't that be great?
You betcha! So that's what the next version of Krakatoa will do (and more). And the more particles you throw at it, the better it will scale. On my workstation, the pure drawing speed of 100 million particles comes close to 17.5 MP/second! The generation of a Matte depth map from 100 million polygons, something that is also performed before each motion blur pass, went down from 32 seconds to 4 (8x4=32, you know?). And we are not even half-done yet. Add to that the loading speed up with PRT Volumes already reported in a previous blog and you will see how this new build is shaping up as the fastest Krakatoa you have ever seen. In the same blog I mentioned it might ship as v1.6.0 - that suspicion turned out true and we already started updating the documentation for this upcoming 1.6.0 build which should be expected sometime before Siggraph.
I am quite excited about these improvements which we haven't even passed to the Beta testers yet as they happened in the last 10 days. All this wouldn't have been possible without our relentless production team which not only pushes Krakatoa to places it has never been to yet, but also constantly bitches about how slow it is! Love you guys! :)
So don't panic, we won't make Krakatoa slower just to sell more network render licenses and Deadline seats to run them, or disable more features so you cannot play with it for free. We understand that a faster Krakatoa is easier to love, and a Krakatoa you love is one you would if not buy yourself, then at least suggest to your employer (the guy with the wallet ;))...
Stay tuned!
Sunday, March 21, 2010
Back To Basics - Krakatoa Best Practices For Fast Iterations.
This morning I received a PM from a CGTalk member with approximately the following content:
Here is my answer, slightly expanded.
Well, we cannot speed up PFlow, so we generally try to work around these issues. Here are some things one could do to speed up work.
1. Save the particles to PRT sequence once, then render as often as you want.
PFlow is slow at updating to a specific frame because it is history-dependent and has to calculate all preceding frames. So if you have 200 frames of animation and want to render frame 190, you have to wait for all 190 frames to update. But if you have already updated frame 190, getting frame 191 should be relatively fast, and this is true for the saving process too. So you set Krakatoa to save PRTs to disk and let it run (overnight?
). Once it is done, you have traded off time for disk space. Now you have Gigabytes of particle data on disk, but you can render each frame in about a second without any preroll! The Krakatoa PRT Loader also lets you tweak the timing/speed of the playback, cull particles, deform the particles if they don't match your vision, bend the particles to follow a spline path, deform by a skinned mesh to create particles moved by a character animation, create and modify particle data channels using MagmaFlow, duplicate the particles by cloning the PRT Loader multiple times and so on. It is a whole new particle data workflow you have there...
2. Partitioning.
You can expand the above approach with some partitioning (saving out several smaller PRT sequences with varying random seeds to produce a denser final result). So you could for example reduce the particle count to something you can live with time-wise, for example 50K, 100K or 500K particles, and partition only one PRT sequence to check out if you like the motion. You could also use the ability of the Iterative Mode of Krakatoa to render a frame at a fraction of the resolution while preserving the density to get a representative frame from few particles that gives an idea what the same frame with 10 times more particles could look like.
If it looks good after it has finished saving, you can let the computer run multiple additional partitions with the same settings to produce more and more particles using the same FumeFX sim (and/or PFlow setup). This way you don't have to wait for the whole quantity to be processed at once and make more iterations until you like the general motion, then let a copy of Max running in the background save 10 or 20 or 50 sequences of the same setup with different random seeds to combine into one dense sequence with millions of particles within the PRT Loader.
3. Use all your CPUs/Cores.
PFlow is single-threaded, but your machine might have more CPUs sitting idle. You can open several copies of 3ds Max + Krakatoa on your workstation (thankfully, both the 3ds Max and the Krakatoa licenses support that), load the same scene in each one of them and let each one partition a sub-range of the total range of PRT sequences. For example, if you have 4 cores, you can open 4 copies of 3ds Max and let the first one partition from 1 to 3, the next one from 4 to 6, the third one from 7 to 9 and the last one from 10 to 12. After about the same time it would take a single copy of Max to create 3 partitions you will have 12 and all cores will be used. I have done this with 8 cores and it works great, assuming memory is not an issue. Saving particles through Krakatoa does not load particle data into memory, only 3ds Max, FumeFX and PFlow will use memory, so chances are this won't be an issue.
4. Render the FumeFX directly.
Krakatoa lets you render every voxel of a FumeFX simulation as a particle. This does not give you the same look as driving PFlow particles through the simulation grid, but for some effects it can be a lot faster.
The drawbacks: The particles don't have velocity so you cannot apply motion blur and you need a very fine-resolution grid to create enough particles, which of course increases the FumeFX simulation time.
Once you have simulated the FumeFX and you like the look, you can just check the ">FumeFX" button in the Krakatoa Main Controls rollout to enable it as source and render away. But you can also use some techniques to increase the particle density. Save the simulation to a single PRT sequence, load in a PRT Loader, add a High-Frequency Noise Modifier to the stack as outlined in this tutorial on our site, check the ">PRT Loader Modifiers" option in the Partitioning Rollout and partition the resulting particle system multiple times to increase the density by jittering the particles around their original positions. The result might look a bit noisy though, but it is a possibility.
Finally, you could just render the FumeFX as voxels, which generally requires less particles to produce the same density. Usually you would need to set the Krakatoa Voxel size to the same value as the FumeFX grid size.
These are some of the typical approaches to speed up the iterative process of creating good looking particles and rendering them in Krakatoa. It does not apply solely to FumeFX and Particle Flow - you can combine this knowledge in other situations regardless of the original source of the particles.
Please add your comments below with your own Best Practices. I will migrate this to the FAQ someday.
May your cores be always busy!
...These days I am studying Krakatoa by myself, it is interesting. Now I can generate particles from FumeFX data. To render Particle Flow, I use 800000 particles, but it is very slow to update PF, takes almost 10 minutes or even more, is it normal? Or did I miss some steps.. My PC is a Dell T5500 workstation, 64bit OS. When PF and Krakatoa work together, how can it be made faster? I can't image using millions of particles...my pc would update PF all day long, haha!For the more advanced users of Krakatoa out there, the following answers might look like deja-vu from the Online Help's Introductory Tutorials and the Studio Daily Introductory Videos posted back in 2007. But I feel it is something that cannot be repeated often enough, and given that nobody ever reads any Help files (I know I don't), a blog post on a site hosted by Google might be more discoverable...
Here is my answer, slightly expanded.
Well, we cannot speed up PFlow, so we generally try to work around these issues. Here are some things one could do to speed up work.
1. Save the particles to PRT sequence once, then render as often as you want.
PFlow is slow at updating to a specific frame because it is history-dependent and has to calculate all preceding frames. So if you have 200 frames of animation and want to render frame 190, you have to wait for all 190 frames to update. But if you have already updated frame 190, getting frame 191 should be relatively fast, and this is true for the saving process too. So you set Krakatoa to save PRTs to disk and let it run (overnight?
). Once it is done, you have traded off time for disk space. Now you have Gigabytes of particle data on disk, but you can render each frame in about a second without any preroll! The Krakatoa PRT Loader also lets you tweak the timing/speed of the playback, cull particles, deform the particles if they don't match your vision, bend the particles to follow a spline path, deform by a skinned mesh to create particles moved by a character animation, create and modify particle data channels using MagmaFlow, duplicate the particles by cloning the PRT Loader multiple times and so on. It is a whole new particle data workflow you have there...2. Partitioning.
You can expand the above approach with some partitioning (saving out several smaller PRT sequences with varying random seeds to produce a denser final result). So you could for example reduce the particle count to something you can live with time-wise, for example 50K, 100K or 500K particles, and partition only one PRT sequence to check out if you like the motion. You could also use the ability of the Iterative Mode of Krakatoa to render a frame at a fraction of the resolution while preserving the density to get a representative frame from few particles that gives an idea what the same frame with 10 times more particles could look like.
If it looks good after it has finished saving, you can let the computer run multiple additional partitions with the same settings to produce more and more particles using the same FumeFX sim (and/or PFlow setup). This way you don't have to wait for the whole quantity to be processed at once and make more iterations until you like the general motion, then let a copy of Max running in the background save 10 or 20 or 50 sequences of the same setup with different random seeds to combine into one dense sequence with millions of particles within the PRT Loader.
3. Use all your CPUs/Cores.
PFlow is single-threaded, but your machine might have more CPUs sitting idle. You can open several copies of 3ds Max + Krakatoa on your workstation (thankfully, both the 3ds Max and the Krakatoa licenses support that), load the same scene in each one of them and let each one partition a sub-range of the total range of PRT sequences. For example, if you have 4 cores, you can open 4 copies of 3ds Max and let the first one partition from 1 to 3, the next one from 4 to 6, the third one from 7 to 9 and the last one from 10 to 12. After about the same time it would take a single copy of Max to create 3 partitions you will have 12 and all cores will be used. I have done this with 8 cores and it works great, assuming memory is not an issue. Saving particles through Krakatoa does not load particle data into memory, only 3ds Max, FumeFX and PFlow will use memory, so chances are this won't be an issue.
4. Render the FumeFX directly.
Krakatoa lets you render every voxel of a FumeFX simulation as a particle. This does not give you the same look as driving PFlow particles through the simulation grid, but for some effects it can be a lot faster.
The drawbacks: The particles don't have velocity so you cannot apply motion blur and you need a very fine-resolution grid to create enough particles, which of course increases the FumeFX simulation time.
Once you have simulated the FumeFX and you like the look, you can just check the ">FumeFX" button in the Krakatoa Main Controls rollout to enable it as source and render away. But you can also use some techniques to increase the particle density. Save the simulation to a single PRT sequence, load in a PRT Loader, add a High-Frequency Noise Modifier to the stack as outlined in this tutorial on our site, check the ">PRT Loader Modifiers" option in the Partitioning Rollout and partition the resulting particle system multiple times to increase the density by jittering the particles around their original positions. The result might look a bit noisy though, but it is a possibility.
Finally, you could just render the FumeFX as voxels, which generally requires less particles to produce the same density. Usually you would need to set the Krakatoa Voxel size to the same value as the FumeFX grid size.
These are some of the typical approaches to speed up the iterative process of creating good looking particles and rendering them in Krakatoa. It does not apply solely to FumeFX and Particle Flow - you can combine this knowledge in other situations regardless of the original source of the particles.
Please add your comments below with your own Best Practices. I will migrate this to the FAQ someday.
May your cores be always busy!
Subscribe to:
Posts (Atom)
Posts
