Archive for June, 2009

Canon VIXIA HF11 Comparison Video

austinTuesday August 5th, the day we went to shoot some HF11 and HV30 video at Auditorium Shores on Riverside Drive in Austin, Texas, was the day that Hurricane Edouard made landfall on the Texas coastline and rapidly moved inward through the state. Needless to say, we were doomed with thoroughly overcast skies, and a light rain started to fall just before we wrapped. We came away with some usable video which served our purposes, but it’s definitely *not* what you would call beautiful footage. It was the middle of the day with a white and gray sky; hardly any color anywhere… the exact opposite of Golden Hour, if there can be such a thing.

Austin Meyers shot the sequences with the HF11 and HV30 camcorders sharing a tripod, which greatly facilitated the ease and speed with which he worked, with the constant threat of an impending torrential downpour hanging over our heads. With the camcorders in close proximity to each other mounted on the same tripod, Austin was able to match camera settings and change frame rates and data rates very quickly.

hf11hv30The sequence involved shooting identical clips with the HV30 recording HDV format video at 60i, 30p and 24p frame rates, with corresponding frame rates on the HF11, which recorded AVCHD format video. Also, the HF11 recorded the same scenes in the new MXP mode at 24 megabits per second, as well as the previously established FXP mode at 17 megabits per second. Both of these bit rates record Full HD video at 1920 x 1080. We also shot a little bit using the HF11’s lesser bit rates, which record anamorphic HD at 1440 x 1080 (as does the HV30 in the HDV format), at twelve, seven and five megabits per second.

The real question at stake is: what difference does the increase in bit rate make, not only between 24Mbps and 17Mbps at Full HD, but the lesser bit rates in 1440 anamorphic as well. The idea here is for you to make your own judgement, by downloading and examining these video clips for yourself. Is the difference dramatic, or is it negligible? Can you detect any visual difference between the MXP and FXP recording modes, or the difference between 1920 square pixels and 1440 anamorphic? You can make a much better analysis by viewing them in full motion on an HDTV instead of a computer monitor.

Video: Fountain Fun at Auditorium Shores in Austin

ausfountWe’ll start with a great codec test: water. Here’s a public fountain near Auditorium Shores in Austin, Texas. We sampled all bit rates of the HF11 as well as 25Mbps HDV from the HV30, at each of the three frame rates (60i, 30p and 24p). Right now we’re still in the process of uploading individual QuickTime .MOV files for each setup, but in the meantime we have a disc image of the entire video sequence from both camcorders available for download now. All of these clips were shot in Tv (shutter priority) mode with a shutter value of 1/100th sec. Exposure and focus were automatic, and the Cloudy setting was used for white balance, as the sky was completely overcast.

Download the HF11 disc image: http://www.dvinfo.net/media/canon/hf11/

Notes on downloading the HF11 disc image: Included in the HF11_DMG folderis a split .dmg disc image file with par2 recovery files. Here’s what to do. First, download all files that end in .00x.

Next, download some or all of the .par2 files (if all the above files download correctly you will only need to download one par2 file to verify) or if you’re feeling lucky, you can bypass the par process entirely.

For Mac OS X folks, make sure you have these programs (they’re free):

MacPAR Deluxe, at http://www.xs4all.nl/~gp/MacPAR_deLuxe/

Split and Concat, at

http://www.xs4all.nl/~loekjehe/Split&Concat/

Then simply open .par2 file with MacPAR Deluxe and it will check and recover all the files and then automatically pass them into Split and Concat, which spit out a .DMG disc image file that you can mount.

For PC folks, similar apps are required that can handle Par2 files as well as an application that can handle split files using .00x extensions (if the par2 app doesn’t). We’ll get some community feedback on this and will revise these notes as soon as possible.

Video: Bit Rate Comparisons, Zooming Sprinkler

sprinkzSince we’ve just shot a fountain, why not maintain a semi-aquarian vibe. For this motion study, we used an ordinary lawn sprinkler to get the motion of the water spray within the frame, and zooming in to get motion of the frame. Each clip goes from roughly just past the focal length midpoint on into full telephoto. The zoom lever was triggered manually, with the zoom speed preset on position number two. The goal is to test the quality of the video compression at each bit rate with movement in the frame and movement of the frame.

These sprinkler zooms were shot in Tv mode at 1/100th sec. with auto exposure settling at or around f/4.8. Instant auto focus was used. These are 60i clips, recorded in all five bit rates: MXP, FXP, XP+, SP and LP.

1. Sprinkler Zoom in 60i at MXP (24Mbps)

2. Sprinkler Zoom in 60i at FXP (17Mbps)

3. Sprinkler Zoom in 60i at XP+ (12Mbps)

4. Sprinkler Zoom in 60i at SP (7Mbps)

5. Sprinkler Zoom in 60i at LP (5 Mbps)

Note the strange file extension of “.m2tx” on each clip. This is a purposefully bogus file extension designed to prevent attempts to open it and play it directly from our server. To view these clips, you’ll need to download them first to your computer. Be sure to re-name the file extension from .m2tx to .m2ts or .mts and open them in any media player that supports these extensions. These video clips are best viewed on an HDTV.

(coming soon: sprinkler zooms in 30p at MXP, FXP, XP+, SP, LP)

(coming soon: sprinkler zooms in 24p at MXP, FXP, XP+, SP, LP)

Video: Bit Rate Comparisons, Panning Sprinkler

sprinkpNext, we continue on with our watery theme. This is the obviously same sprinkler setup as above, with about the same amount of water pressure going through it, and the HF11 camcorder at approximately the same distance. This time we’re panning from side to side. It’s not intended to give you motion sickness; it’s just another way of testing the robustness of the compression used at each bit rate. If you’re watching these on a large HDTV (and hopefully you are), then you’re advised to sit down, lest you might reach for the Dramamine tablets.

These sprinkler pans were shot in Tv mode at 1/500th sec. with auto exposure hovering around f/4.0. Normal auto focus was used. These are 60i clips, recorded in all five bit rates: MXP, FXP, XP+, SP and LP.

1. Sprinkler Pan in 60i at MXP (24Mbps)

2. Sprinkler Pan in 60i at FXP (17Mbps)

3. Sprinkler Pan in 60i at XP+ (12Mbps)

4. Sprinkler Pan in 60i at SP (7Mbps)

5. Sprinkler Pan in 60i at LP (5 Mbps)

And once again, note that every clip has the strange file extension of “.m2tx” and this is intentional. This is a purposefully bogus file extension designed to prevent attempts to open it and play it directly from our server. To view these clips, you’ll need to download them first to your computer. Be sure to re-name the file extension from .m2tx to .m2ts or .mts and open them in any media player that supports these extensions. These video clips are best viewed on an HDTV.

(coming soon: sprinkler pans in 30p at MXP, FXP, XP+, SP, LP)

(coming soon: sprinkler pans in 24p at MXP, FXP, XP+, SP, LP)

Stills: HF11 and the DSC Labs ChromaDuMonde Chart

dscchromaI have the 24Mbps video versions of these images, but they’re about twentyfive megabytes apiece and each one is only a few seconds long. However, these still frames should work fine for analyzing color accuracy with a vectorscope, because they were processed in the video colorspace. In other words, these aren’t photographs that were taken in photo camera mode, which are processed differently than video — these really are frame grabs from video, acquired using the Still Image Recording function in the camera’s video recording mode. Click the link to open a full-size 1920 x 1080 image. They were shot in the “P” Program Autoexposure mode, which the camera established at f/4.0 and 1/150th sec. The white balance settings used were Auto, Daylight, Shade, Cloudy and Manual. The images are about one megabyte each.

1. DSC Labs ChromaDuMonde, Auto WB

2. DSC Labs ChromaDuMonde, Daylight WB

3. DSC Labs ChromaDuMonde, Shade WB

4. DSC Labs ChromaDuMonde, Cloudy WB

5. DSC Labs ChromaDuMonde, Manual WB

Archives

You are currently browsing the Rebel Filmmaker blog archives for June, 2009.

Bookmarks