There are 3 ways to convert 60i/30p to 24p. Each one has its ups and downs. None is really ideal for all cases, so if you really need real 24p, get a real 24p camera. In detail:

1. Resampling (or motion estimation)
This is the default Sony Vegas mode when your project properties has a specific source frame rate and you export in another, or when you slow-down footage. There is a huge amount of ghosting introduced when resampling is ON. I personally always make sure that all the clips in my timeline are set to “disable resample” before I export. I suggest you do too. Vegas’ resampling is one ugly algorithm, avoid. Ghosting:

2. Drop frames
This method will drop 6 frames per second in order to convert from 30 to 24 fps. On Sony Vegas you do that with the “disable resample” option mentioned above. Other editors, use this method by default (it’s the simplest to implement). The problem of this method is that the video it produces is jerky. You see, the camera didn’t record 24 frames at 24 even times a second, but 30, and so there will be “gaps” in the motion of the video. These gaps will show up as jerky video. The good news is that many viewers never notice, the bad news is that this will still not visually feel like real 24p.

3. 25% slow down
This is my preferred method. Basically, by slowing down the playback rate by 0.800, we essentially emulate the 24p capturing process perfectly. There is no jerkiness, everything gets a smooth, dreamy look, which is very cinematic. And no, the vast majority of the viewers will never realize that the footage was slow-motion’ed! It is the perfect way to shoot music videos (using this trick), artistic/abstract pieces, and anything else that doesn’t require synchronized speech. See, this method has a drawback: you can’t resample audio with the same effectiveness as you do for video, it will sound real slow and bad (even after correcting the pitch). So basically, this method is out from the moment you want to use the audio captured along the video. You can only add music/narration during post processing, but no sync’ed speech. This method takes out the ability of shooting short movies where people have to talk on camera, but it’s good for everything else.

Here are the playback rates you need to use for the 3rd method:
60i (via interpolated de-interlacing), or 29.97 fps to 23.976 fps: 0.800
60i (via interpolated de-interlacing), or 29.97 fps to 24.00 fps: 0.801
30.00 fps to 23.976 fps: 0.799
30.00 fps to 24.00 fps: 0.800
Of course, always disable resample just before you export, on all your clips in the timeline.

I am shooting a wedding for a friend this weekend, and along enough tapes, a second battery, and two additional HD digicams (Panasonic FX150 and Canon SX200 IS) sitting on tripods, the following Canon HV20 camcorder will be the main rig that I will be using.

JBQ will be shooting still pictures with his Canon 5D, and he will be taking care one of the two digicams (video recording must be restarted every 15 minutes, because of FAT32 limitations). I just need to find a second person at the wedding (possibly a kid) to take care of the second digicam.

There is currently no stabilization plugin for Vegas Platinum since the only third party plugin that used to be available from BorisFX was discontinued a few weeks ago, while Deshaker/Mercalli only work with the Pro version of Vegas. So I decided to write this little Deshaker/VirtualDub tutorial to show you how to stabilize your footage, especially if you are running Vegas Platinum.

Update: Below, I’m suggesting the Lagarith codec, but you can try instead the new Matrox MPEG-2 AVI intermediate codec.

Vegas Pro

For Vegas Pro you simply have to install “New Deshaker” script. Install the New Deshaker .msi or .zip file (choose one of the two, files are to be installed on the Vegas Pro’s “Scripts” folder if you choose the .zip file), and the VirtualDub package also from the same page.

Then, install the 32bit version of the Lagarith codec from here. If you are running a 64bit version of Vegas, install both the 32bit and the 64bit version of Lagarith, otherwise, just the 32bit version (regardless if your Windows is 32bit or 64bit).

Load Vegas, load the file you want to stabilize, setup the right project properties (this is important, follow step 1 from this tutorial). Place the file in the timeline. Click “Render As” from the main menu, select the AVI filetype, and then click “custom” to make everything look similar to this (depends on what footage you used. Also, ignore the project properties step, since we did that above). Then, give a new name to that tweaked template, e.g. “Lagarith”, and then click the save icon in that “Custom Template” dialog. From now on, you can reuse that template directly.

Discard the “Render As” screens, save the whole Vegas project and give it a name. Now select the clip in the timeline, and select the New Deshaker option from Tools/Scripting. In the new window that’s loading, make sure the paths to VirtualDub/Deshaker are correct (for 64bit operating systems you might need to tweak them), set a “render to” path (e.g. c:\videos\myVegasProject\temp\), and select the “Lagarith” template we just created above in the “Outbound Template”. Then, in the “VirtualDub Compressor” option, delete the “0” there, and type: “lags” (without quotes). Then, “Save settings” in the bottom of the New Deshaker window, and then select “Start”.

The script will now start stabilizing your video, it might take 3 to 8 minutes to stabilize just 10 seconds of footage depending on the speed of your CPU, so use this ability wisely. After it’s all done, the new stabilized video will be inserted as a new “Take” on top of the existing clip in the timeline. By pressing the “T” key on the keyboard, while the clip is selected in the timeline, is going to interchange the take between the original file and the stabilized one.

Please note that the new stabilized take, which uses the Lagarith codec, will be slow to edit, this is to be expected. Unfortunately, there is no freeware AVI intermediate lossless codec that’s fast to decode, so Lagarith is your best option. The filesize is going to be big as well (~2 GBs per minute), as it’s true for any intermediate lossless codec.

Also, if you find that a Vegas Pro process still running on your Windows after you have closed down Vegas, then you must kill that process using the Windows task manager (New Deshaker doesn’t always clean up after itself).

Finally, in the New Deshaker dialog, you can create a new template instead of the “Default” one, to optimize the stabilization even more. For example, in the “Pass 1” tab you can change “Half” to “Full”, and “Every 4th” to “All”. Doing these changes will add more stabilization into the video, but it will also make the calculation process 4-5 times slower (and don’t forget that it’s already slow).

Vegas Platinum

Install this. Then, install the 32bit version of the Lagarith codec from here.

Load Vegas, load the file you want to stabilize, setup the right project properties (this is important, follow step 1 from this tutorial). Place the file in the timeline. Click “Render As” from the main menu, select the AVI filetype, and then click “custom” to make everything look similar to this (depends on what footage you used. Also, ignore the project properties step, since we did that above). Then, give a new name to that tweaked template, e.g. “Lagarith”, and then click the save icon in that “Custom Template” dialog. From now on, you can reuse that template directly. Now, from the main “Render As” dialog, render out the file.

Load the VirtualDub.exe application from its “Program Files” folder (navigate there with Windows Explorer). Select “File”, and “Open Video File”. Load in it the .avi file you just rendered above. *If* your VirtualDub is freezing while loading any AVI file and it shows just black screens, then you need to go to its Preferences panel and select OpenGL for output, and VSync (at least, this fixed the issue for me). Anyways, when the AVI file is loaded on VirtualDub, click on “Video” from the main menu, then “Compression”, and from the long list that pops up, find the “Lagarith lossless codec” and select it. Click “Ok” to that dialog.

Then, from “Video” again, select “Filters”. Click “Add”, scroll and select “Deshaker 2.4”, then click “ok”. Select the “Deshaker” entry in the Filters dialog and click “Configure”. Set the “Source Pixel Aspect” the same way as it’s on your Vegas’ project properties (e.g. for HDV it’s 1.333, for most AVCHD it’s 1.0). For “Video Type” select “Interlaced, upper field first”, provided your video is HD and was shot interlaced (if unsure, then again consult Vegas’ Project Properties as you are supposed to have the right properties). Finally, click on the big “Pass 1” button. Then, click “ok” and “ok” again.

Now, click the “play” button on the VirtualDub icon bar on the bottom of its window. Make sure you click the play button that has the little “o” in it, not the one with the “i”. It will now apply the Pass 1 of Deshaker, have patience.

When it’s done, select again Video/Filters, and select the Deshaker plugin, and then click “Configure” again. Then, select the “Pass 2” big button. Click “ok” twice again. Now, click on “File”, and “Save as AVI”, give it a filename. It will now export the stabilized video in the Lagarith AVI format.

The new stabilized file is now ready to be edited on Vegas in your main project, so bring it in! *If* Vegas recognizes the stabilized file as 4:3 instead of widescreen, then while it is still in the “media bin” holding area of Vegas, right click on it, “Properties”, and from the “Video” tab set its aspect ratio to 1.3333 instead of 1.000. You can even tell Vegas to remember that setup for that specific video format, by clicking the little “save” icon in that dialog.

Please note that the new stabilized version, which uses the Lagarith codec, will be slow to edit, this is to be expected. Unfortunately, there is no freeware AVI intermediate lossless codec that’s fast to decode, so Lagarith is your best option. The filesize is going to be big as well (~2 GBs per minute), as it’s true for any intermediate lossless codec.

Finally, in the Deshaker configuration dialog inside VirtualDub, you can optimize the stabilization even more. For example, in the “Pass 1” tab you can change “Half” to “Full”, and “Every 4th” to “All”. Doing these changes will add more stabilization into the video, but it will also make the calculation process 4-5 times slower (don’t forget that it’s already slow).

Some more color grading, some extreme, some not. Lighthouses was my subject tonight. I used mostly Magic Bullet, Bump Map, Color Corrector, Unsharpen Mask, and a variety of Pixelan’s Vegas plugins.

Before:

Picture by chjab, licensed under the Creative Commons “Attribution” 2.0 license.
After:

Before:

Picture by clapon, licensed under the Creative Commons “Attribution/Share-Alike” 2.0 license.
After:

Before:

Picture by Bocian & Tusia, licensed under the Creative Commons “Attribution” 2.0 license.
After:

Before:

Picture by Jake Wellington, licensed under the Creative Commons “Attribution” 2.0 license.
After:

After Adam and Thom, I will be putting a break on my Apple purchases too. I have already promised days ago one more Apple iPhone review to a third party store, so when I am done with that too, it’s over for me as well.

Apple is quickly becoming the new Microsoft in the eyes of the people. The funny thing is though, that Apple was always like that, it’s just that people outside of the Bay Area didn’t know all of the juicy details most of us residents know. You see, Silicon Valley is a small place. You will be surprised how small it is. Word goes out easily. So while I might not have been blogging or reporting much about small tidbits that I happened to hear over the years (in order to protect my sources), the truth is, Apple never had a good name as a workplace/business in the area. Not before Steve Jobs came back as a CEO, and certainly not after.

But I guess, the cat is out of the bag now.

Between the crude iPhone application authorization process, the no-background processes allowed, to selling only locked phones (something that’s pretty illegal in other countries and I hope it would be here too), it makes me loathe what Apple is doing. They have created the best smartphone experience, they started the true smartphone revolution with the iPhone, but at the same time they try to limit progress in other areas.

It’s not a coincidence that the un-approved, and previously approved by Phil Schiller himself, iPhone apps were ALL Google Voice-related (I think there were 2-3 Voice-related apps un-approved, plus the Google Voice official app that was unauthorized from the get-go). The hard work and sweat of these developers, all went to waste. It’s more than obvious here that AT&T is behind this plot. You see, Apple has nothing to lose with Google Voice, if anything, Apple has everything to gain from it (it makes their phone more useful)! But AT&T is the one who has everything to lose. Google Voice allows for free US-bound calls and dirt cheap international calls (just $0.02 per minute for US to France/Greece), which of course, puts AT&T’s business at risk.

What Google is doing with Google Voice is nothing but progress. They have the bandwidth, so they go with it. AT&T on the other hand, is nothing but a new RIAA/MPAA, scared of the new realities that technology brings! They can’t, or they don’t want to, change their business and/or technologies, and so they fight against the new kids on the block, who use technology in a more flexible way.

Put that in addition to what AT&T did to my iPhone last month: they cut off my EDGE support. My iPhone is LOCKED to AT&T, and it is NOT jailbroken. It is as vanilla as it goes. The only difference here is that I didn’t buy the iPhone from AT&T at the time, as I only needed to use their PayAsYouGo plan (since I do not do more than 4-5 calls per month).

Think about it. AT&T blocks NO OTHER cellphone-maker for that plan! They single-out the non-AT&T-bought iPhones, like they are a plague, EVEN if they might be as vanilla as they get. THIS ALONE can be used for a class action lawsuit. I had no plans contacting EFF about it, but be sure I will do so now, after the latest Google Voice fiasco. There are too many things to hold a grudge now, I am afraid.

In the meantime, I have emailed Apple with feedback about their practices, and I suggest you do so too.

Update: Kroc wrote on Twitter that “I think you peeps should create a charter that defines “iPhone fixed” and publish it“.

Here’s my list, in this order, for Apple:
1. No more locked phones. Phones should also be able to be purchased at full price, with no ties to any carrier. Subsidized phones with a contract should also be unlocked.
2. Authorization denial of iPhone/iPod apps should be restricted to malware/spyware/buggy apps, and to illegal apps (e.g. a Nazi-related app). All other apps should be allowed to go through to the Apple Store. Sexual-related apps should be allowed, but with age verification or warning.
3. Allow background processes on the iPhone/iPod, as long as they are don’t seem to be compromising the system (e.g. battery life, system software, towers). The iPhone is not a real smartphone without background processes.

For AT&T (and ANY other carrier):
1. Stop dictating to manufacturers what software they can put in there and what they can’t. It’s not your job. It only becomes your job if your towers are compromised. Otherwise, SHUT IT.
2. No more locked phones. Period. You can still subsidize phones, but they have to be unlocked.
3. Allow EDGE/3G/GPRS for all phones. Artificially limiting the non-contract-bound Blackberries and iPhones, is unacceptable, even if I am willing to pay up your crazy prices ($10 per 1 MB of data transfered)!
4. Allow “PayAsYouGo” calls from Europe and other places within US (e.g. the Ukiah area in CA, USA). It’s *unacceptable* to not be able to use my phone when I am on vacations (I am blocked from calling out in these areas), even if I am willing to pay up your crazy ass prices!

Finally, if Apple can’t design protection to open the app CPU without compromising the radio CPU, they need better engineers. The notion that the towers aren’t designed to deal with faulty clients is just bullshit. Who would design a client-server system where the server trusts the client?

Panasonic announced yesterday the FZ38, an 18x super-zoom camera (called FZ35 in US). Not really the brightest idea for a digicam purchase. However, this camera has other features that can prove very useful to people who need 24p with manual controls for cheap ($400).

See, this camera is the first consumer non-DSLR digicam that offers manual controls in video mode! It has both shutter speed and aperture controls! At this point, I can only assume that when in this manual video mode, the exposure will stop jumping left and right as it does currently with other Panasonic digicams.

Then, there’s the frame rate thing. According to DPReview, the camera can do 25p and 30p (depending which model used, European or US), at 17 mbps AVCHD-Lite. However, it saves the videos in a really bad way, and makes editors *think* that it has recorded in 50p and 60p, while in fact it just has duplicated the frames. On Sony Vegas you have to specifically tell it to use either 25 fps or 30 fps (depending if you used the European or the US model) in order to avoid the duplicated frames. Sample 25p/50p .mts file here (bottom of the page).

So in a scenario where you shot in 25p (with 1/50th shutter speed), you edit as such, and at the end you re-time the video to become 24p (if desired). This will produce a very filmic motion look, as close as it gets with a digicam. Yes, there are other cheap digicams that do 24p right out of the box, but they don’t offer manual shutter speed control, which is an important ingredient in the quest to get the filmic motion look.

Additionally, color/saturation/contrast/brightness controls are offered in video mode, as well as manual white balance. In conclusion:

Pros:
– Shutter speed manual control
– Aperture manual control
– 25p can easily be converted to 24p
– Color adjustments
– Manual white balance
– Manual focusing in addition to auto focusing
– By using a CCD sensor won’t produce wobbly videos

Cons:
– Small sensor (1/2.33″)
– Its doubling of frame rate is stupid and unnecessary
– Not a fast lens (meaning, less background blur than the HV20)
– To get the 25p recording you need to buy the European (FZ38) version.
– We are still not sure if exposure will continue jumping even when in manual control mode (there was no word about ISO/gain control you see)
– Its 17 mbps bitrate is much lower than Canon’s 24 mbps digicams in 720p

Before:

Picture by Alex Witherspoon, licensed under the Creative Commons “Attribution” 2.0 license.

After:

Used Sony Vegas’ color corrector to fix the white balance, then used a modified “bleach bypass” Magic Bullet look, and a bit of unsharpen mask to give it a more filmic look.

Geeks.com, sent over one of their popular computer parts products, video cards, the nVidia GeForce 8800GT (512 MB DDR3, PCIe, PureVideo2 HD support, GL 2.0, DirectX10, HDCP). This is a test on video playback performance with Vista-default drivers (Vista 64bit, SP2), and nVidia accelerated drivers (latest stable, v190).

I tested a Canon 5D Mark-II file, since it’s a heavy format: MOV h.264 High Profile, 40 mbps, no audio. File was loaded full screen in a 1:1 screen (1920×1080) at 32bit color, using various decoders and media players. Then, the frame rate and CPU usage was measured. I used a video file with a lot of movement to visually figure out if VLC was playing the file in real time or not (from the players I tried, it was the only one that didn’t have a way to show fps performance). The rest of the players had a way to get actual concrete numbers. Results below:

The CoreAVC Pro CUDA-accelerated version had of course the best result with just 3% CPU utilization (the Vista default drivers had no CUDA support). When CUDA was turned off, there was still a small speed up with the newer, non-Vista, drivers. The rest of the decoders also had it easier either with better frame rate, or with less CPU utilization. If they didn’t do better in terms of frame rate was mostly because of multi-threaded issues, as these decoders are written in legacy styled code (JBQ and I still joke sometimes how even today’s programmers can’t get multi-threading). The only decoder from the ones I tested that was actually multi-threaded was CoreAVC’s. These guys rock.

Please note that I used a speed up option for VLC to get real time decoding with it. By default, VLC doesn’t do real time on the 5D files, not even in this Quad Core 2.4 Ghz DELL PC I used for the test.

The moral of the story is:
– Use graphics cards that have a fast memory bus. Since 2D acceleration tapped by generic non-Purevideo decoders is mostly bandwidth-bounded, get cards that don’t cut costs by using slow memory or buses.
– Don’t leave your PC with the default XP/Vista/Win7 drivers. Upgrade to the latest stable version from your manufacturer’s web site.
– When possible, use CoreAVC Pro as your default decoder on media players/editors (Vegas won’t support it unfortunately, since it doesn’t support DirectShow decoders — but Premiere might).
– Prefer nVidia over ATi. nVidia’s PureVideo architecture is better supported by decoders, be it CoreAVC or Adobe’s CS4.
– Don’t ever opt for an Intel integrated card, unless you are really short on money.