XF100 first impressions

[Jack Frost]

Hi zagnut,

thanks for the response. I did a search and found out that yes, it's 25Megabits per second. You are pretty good !
http://www.camcorderinfo.com/content...iew/Format.htm

thank you for the info below, very good to keep handy.

Can you tell me about AVCHD, do Canon SLRs use AVCHD ? the files I produce with my Canon T2i are in the .mov format, and I am sure AVCHD files are with the .MTS format.

I'm not positive, since I've never dealt with DV or HDV, but I believe the bitrate is 25Mb/s or about 3MB/s.

There are camcorders that can record video much better than a DSLR, bigger sensors, 3 sensors, etc, but they cost a lot of money. I also know some Panasonic cameras record at 100Mb/s (about 12MB/s). $$$$+

HDV cameras, like your HV20, encode footage using h.262 MPEG-2. While the DSLR's and camcorders mentioned earlier encode using h.264 MPEG-4. Bit for bit, h.264 is more efficient than MPEG-2. Meaning that at the same data rate, H.264 usually offers better quality, if everything else were equal and the same. However, h.264 requires much more in terms of processing power. Long story short, MPEG2 at 25 Mbps should look about as good as anything you get from a broadcast HDTV signal or an H.264 file at a lower data rate (say ~10-12 Mbps). MPEG-2 is also easier to edit / work with and faster to encode when compared to h.264.

[Jack Frost]

for anyone interested, I found a link to the Canon XA10 camcorder user manual in .pdf format:
http://xa10forums.com/Canon_XA10_Manual.pdf

and yes, there's a new forum specifically for the Canon XA10.
http://xa10forums.com/

[zagnut]

Hi zagnut,

Can you tell me about AVCHD, do Canon SLRs use AVCHD ? the files I produce with my Canon T2i are in the .mov format, and I am sure AVCHD files are with the .MTS format.

Both the .mov from the Canon DSLR's and AVCHD .mts files are h.264. They use the same type of compression. They are just packaged differently.

[Jack Frost]

Last night I was looking for Canon XF100 video on YouTube and came across a user that posted a link to a lot of raw files from his Canon XF100.

If anyone wants to play with these, here they are:

There are 20 raw files available ( .MXF files )
http://www.mediafire.com/?vrp637b9gdc0m

I found the links here at the bottom of this video:
http://www.youtube.com/watch?v=6Qmhku1YnS8

Youtube user: http://www.youtube.com/user/unimpacked

[peterarcson]

Allan Roberts PAL XF100 tests:
http://thebrownings.name/WHP034/pdf/..._XF100-105.pdf
"2.8 Conclusion
This camera performs reasonably well at HD, for such a small-image format with a single sensor. Resolution
is maintained up to 1280x720, but contains inevitable spatial aliases due to the use of a Bayer-patterned
sensor. Detail controls work reasonably well, but not as well as they would have done with 3 sensors. Noise
levels are typical for a ⅓” single sensor, and sensitivity is as expected. Operating the camera at significantly
lower gain will reduce the noise level without sacrificing significant sensitivity.
The integral lens has a maximum aperture of F/1.8, normal for a small camera, and there was no perceptible
loss of resolution through iris diffraction until the lens was stopped down to F/11, at which point significant
resolution was being lost (iris diffraction would normally be expected to start at about F/5.6, but the spatial
aliases effectively disguise the effects of diffraction). Thus, the camera has a useful aperture range from
F/1.8 to about F/10. This, together with the in-built neutral density filters (up to 3 stops), means that the
camera has a good exposure control range.
Performance at 720p is acceptable, since it appears to have been derived directly from the sensor, rather than
from 1920x1080 signals. 720p performance can be improved significantly with the use of the noise reducer,
which also reduces video noise to a respectable level.

[BIG SKILLY]

im picking up my xf-100 this weekend...1 question...what type of system requirements for mac do i need??...cant wait to get it i will show footage once i do

[Jack Frost]

Big Skilly, are you watching the thread on the XF-100 over here ?
http://www.avsforum.com/avs-vb/showt...2#post20416602

[BIG SKILLY]

Big Skilly, are you watching the thread on the XF-100 over here ?
http://www.avsforum.com/avs-vb/showt...2#post20416602

Thanks i did...read all 60 pages and nothing about the xf-100...only thing they spoke on was the xha1...

[cgbier]

Ah, another 720p sensor. That'd be fine with me, as my end product is either DVD or 720p web anyways.

If I just could find the money.

[peterarcson]

For Apple/Mac you will need FCP 6 or 7. There is a plugin either on the disc or download from Canon web site. You can edit native, don't have to transcode. FCE and iMovie will not handle the files.
XF100 is not 720 sensor, it is a full 1080 sensor, single chip 1/3".

[ejolson]

The imager of the FX100 has 2371040 photo sites arranged in a Bayer pattern. Assuming a 4:2:2 color space

720p video has 1280 x 720 Y + 640 x 720 U + 640 x 720 V = 1843200 distinct values per frame.
1080p video has 1920 x 1080 Y + 960 x 1080 U + 960 x 1080 V = 4147200 distinct values per frame.

Since there is no way to get 4147200 distinct values from 2371040 photo sites, the FX100 does not produce full resolution 1080p video with a 4:2:2 color space. However, 720p is possible.

[peterarcson]

I am sure you'd know better then Allan Roberts, who posted the report.
"Fig.1 shows a single quadrant of one pattern; for this
exposure, the camera detail enhancement was turned
down to minimum level (-10) which presumably
means no correction, so this is probably the native
performance of the camera. There are strong diagonal
aliases, and coloured horizontal and vertical aliases,
which is a clear indication that the sensor has a Bayer-
pattern structure. Also, the fact that the coloured
aliases are centred on the extremes of the pattern
proves that the sensor is 1920x1080, confirming the
specification in the manual"
"In a camera with a single Bayer-patterned sensor, the red and blue patterns each
have half vertical and horizontal resolution, because the pattern of pixels alternates
between …RGRGRGRG… and …GBGBGBGB… line by line, thus red pixels
appear in alternate columns and lines as do blue pixels. Therefore the clean
resolution limits for red and blue are half that of the sensor. The green pixels form
a quincunx array (Domino-5 pattern), which is best understood as an array of
diagonal lines rather than horizontal and vertical. Thus, the green resolution is
rotated by 45º, and therefore reduced by 1/√2. This results in the green resolution
achieving full horizontal resolution when there is no vertical resolution, and full
vertical resolution when there is no horizontal resolution, a diamond shape rather than square with the
diagonals reaching the sensor’s horizontal and vertical extremes. The diamond nature of this resolution is
clearly visible in Figure 1, where the demarcation between wanted and aliased resolution is a diagonal line
from the horizontal extreme to the vertical. The coloured aliases around 1920 horizontal and 1080 vertical
are due to the reduced nature of the red and blue resolutions.
The only way to get good resolution from a Bayer-patterned sensor is to have more pixels, typically
2880x1620 would be enough to deliver full resolution at 1920x1080Figure 2 Bayer pattern "
"The internal down-conversion to 1280x720 performs as expected, frequencies outside the resolution limits of
1280x720 are aliased in the image. Better down-conversion is possible using a separate conversion process,
either in software or hardware. However, the noise reduction filtering can be used as an alternative, as will
be shown later. "
In short I am trusting his opinion, not yours.

[ejolson]

I am sure you'd know better then Allan Roberts, who posted the report...In short I am trusting his opinion, not yours.

Dear Peter, I'm in full agreement with Allan. I only explain the mathematics behind his statement that "typically 2880x1620 is enough to deliver full resolution 1920x1080" and "internal down-conversion to 1280x720 performs as expected." I think the XF100 is a really nice camera and am looking forward to hearing more about it.

[BIG SKILLY]

For Apple/Mac you will need FCP 6 or 7. There is a plugin either on the disc or download from Canon web site. You can edit native, don't have to transcode. FCE and iMovie will not handle the files.
XF100 is not 720 sensor, it is a full 1080 sensor, single chip 1/3".

Thanks for the info...i have fce now i have to get fcp...anyone has a free download lol...the fx100 has a wide lens?...

[BIG SKILLY]

also the camera doesnt support tiger for macs?

[peterarcson]

also the camera doesnt support tiger for macs?

Sorry, but what are you talking about?
Yes, XF100 has wide angle lens- 30mm in 35mm equivalence.

[cgbier]

Skilly, did you check if the Canon transfer software works with Tiger? Snow Leopard is only 30 bucks (if you have an Intel machine), and there is plenty FCS2 for cheap on ebay.

[peterarcson]

Tiger is OS, FCP is a program. FCP can run on any OS above 10.4. I am on 10.5 running FCP 6.0.6 and the plugin for L&T works just fine. In June Apple is releasing the new FCPX, which is going to cost $299, so if you don't own it yet just wait till then. I am still keeping FCP6, because I own the whole Magic Bullet, and I don't want to upgrade that. Most likely I will wait with the purchase to see how it works, as I have a working edit station and I don't want to change anything else.
To answer your question Tiger has no bearing on how Canon plugin works, as it works in FCP program and has nothing to do with OS itself.