Canon XA10 - anyone seen footage?

Also should mention the level of control over the image on the XA10/HFG10 will not be anywhere near what you have on the XL/XH/XF lines. So another thing to really consider if trying to move to these new cams.

I flying out to Vegas for CES shortly, so I will investigate this myself when I am there. I'll try to take video and pics...
 
mgalvan said:
Also should mention the level of control over the image on the XA10/HFG10 will not be anywhere near what you have on the XL/XH/XF lines. So another thing to really consider if trying to move to these new cams.

I flying out to Vegas for CES shortly, so I will investigate this myself when I am there. I'll try to take video and pics...
Click to expand...

michael-

thanks for the kind offer.

i look forward to your insights.

be well

rob
 
mgalvan said:
Also should mention the level of control over the image on the XA10/HFG10 will not be anywhere near what you have on the XL/XH/XF lines. So another thing to really consider if trying to move to these new cams.

I flying out to Vegas for CES shortly, so I will investigate this myself when I am there. I'll try to take video and pics...
Click to expand...

Much appreciated!
 
Nektonic said:
If it is comparable in terms of image, it seems that maybe one of those XA10 cameras might be a good B-cam for an XF100 if you were needing a two cam setup for ENG stuff or event production.
Click to expand...
I think the main advantage of the HF100 over the XA10, image-wise, is the MPEG-2 CODEC. The only real win is 4:2:2 color, unless you have a need for an MPEG-2 workflow, can't or don't want to handle AVC. Otherwise, they're using the same single-chip CMOS sensor.

I'd really like to see some actual results... I'm skeptical. I suspected that the move here was simply that Canon developed this new sensor primarily for the XF300 series (3-chippers), loved it, and decided to use it in other models. But the single-chip low-res sensor... I did that whole thing with my Sony HVR-A1 for the last 4-5 years. You will get more sensitivity, all things being equal, than going to the larger pixel-count sensors they've been using in most consumer models recently. But you lose the good color and overall sharpness, since now you have Bayer interpolation to deal with. And while the 4:2:2 would lead to better chromakeying, the single chipper has to erase that possibility.

I was expecting something profound, and I guess the tests will materialize. But Canon's own site claims this is 25% more sensitive than their current 8.5Mpixel Vixia sensors... that's kind of a disappointment. The HF S30, for example, has a slightly larger single chip (1/2.6" vs. 1/3"), but you'd expect a technology improvement as well, plus the larger sensor sites, to add up to more than 25%... that's like what, 1/4 f-stop. If that's a real number, they could do better just bumping the lens from f1.8 to f1.6. Or am I missing something?

Nektonic said:
The XF100 seems to be the closest thing to a tapeless XHA1 in terms of ergonomics and price. Then again, the XF100 is the same price as an HMC150.
Click to expand...
And it's hard to imagine buying a single-chip camera over a 3-chipper. I'm certain Canon's new 1/3" CMOS chip is a technological wonder, but is it really 3x as sensitive as the CCDs in the HMC150... that's really the question you wind up asking.

Nektonic said:
If I were still going the traditional 1/3 chip video camera route, I'd look at the XF100 and maybe the XA10 as replacements for my XH-A1 and HV20. Personally though, I'm more interested in Canon's DSLR line, the Panasonic AF100, GH2, and that Sony NXCam 35 that will be coming towards the middle of the year.

I agree on the tape thing though. I am ready to go fully tapeless in 2011.
Click to expand...

I did the tapeless thing in 2010. I had been playing around with AVC cameras and tapeless for a few years. I bought one of the Hitachi Blu-ray camcorders at a fire sale price... really very much a consumer model, even at the original $1500 or so price. But it was pretty clear that AVC was not yet ready (in theory, it's twice as efficient for the same visual quality, but that's only true with an advanced enough encoder... which meant waiting for a fast enough video DSP that could crunch AVC for the 2-3W power a camcorder is likely to have available). I tried again with the low-end Sanyo Xacti FH1, and that could actually compare to my Sony HVR-A1 video... softer lens, but oddly a tad better in low light.

So I upgraded to the HMC40, and added a TM700 as a "B" camera. I'm totally happy with those... but yeah, next move is to add a video DSLR. That's a little frustrating, too... given that practically every model does 1080/30p, 1080/24p, and 720/60p, usually from a 10-20Mpixel sensor, it's only a matter of time until we get 1080/60p and maybe some 3K or 4K modes. I've convinced myself that the day I upgrade, someone will announced that 4K model :-)

The DSLR seems to me the way to deal with low light... rather than comparing 25% differences in light sensitivity between single sensors, or maybe a 300% difference if you talk 3-chip vs. 1-chip, you're throwing a 1400% (four-thirds), 2400% (APS), or 5000% (35mm) sensor at the problem. And of course, with the higher pixel count, there's plenty of pixel bucketing, so no worries about color fringing (sure, you have a new set of worries, like pronounced "jello" effect, etc).
 
Last edited:
hazydave - check out the side by side comparisons on Slashcam with the XF100. It looks to outperform every 3-chip CCD model in the database (HMC151, XH A1, HPX170, etc). It even gives a lot of 3-chip CMOS cams a run for their money, and arguably trounces some of them as well (in terms of color, resolution & sensitivity) - Sony's NX5u, most notably.

There's no data yet on the XA10, however I expect it to do well, but just fall short of XF100 performance - mostly due to codec constraints.

http://camcorder-test.slashcam.com/campair-EN.shtml
 
Yeah, I saw that... I'll admit, I'm suspicious. Particularly because their data, and particularly image capture, also shows the XF100 better than the XF300. That can't actually be possible... three of the same senors, versus the one.

I'm not shocked that overall color would look better than many 3-chippers given enough light, based on state-of-the-art CMOS and 4:2:2 encoding.... my experience with the HVR-A1 suggests that the higher color resolution of a good CMOS chip can help, and that the Bayer interpolation does really affect color perception (you have the same amount of color samples prior to MPEG or AVC encoding), just the accuracy (you will have interpolation errors on some material). I'd like a little more information on this model, too... could be their image processing can account for boundaries in ways impossible before this.

But why the 12lux XF100 looks better than the 12lux XF300... something ain't right.
 
I pretty much have decided that Slashcam is a work of fiction... who are these guys, and why does anyone trust their results. I worked in Germany on and off for six years, even had a startup there (Metabox AG)... I tend to believe the typical German "ruthless efficiency" makes them good engineers. But the results show here just simply defy credibility... and not just because they have the XF100 beating the XF300 and XF305.

I mean, play with it a little. I found out that not only is the TM700 delivering a better 12lux image (better color, better contrast, and only slightly worse noise) than the HMC151, but even the frackin' Panasonic SD9 is competitive here! Have you ever used an SD9 in a room that's not lit with sunlight or 6000W worth of photo lights? My daughter has one.... I bought it for $200 a few years back (she's in a Comminucations specialy in High School). Great value, nice video in bright light, native 24p mode... but it's a 1/6" 3-chipper, abysmal in low light. And yet, sure looks better -- a little grainier, but better contrast and color -- than the HMC151 on Slashcam. Go to the color graph, and they're nearly identical, the SD9 and the HMC151. That SD9 is even nearly as low-grain as the TM700, and the white balance is much better, in their image capture.

And do they ever bother with white balance on those low-light shots, or just trust the camera's auto-mode. I mean, sure, that's a thing consumers do, but would anyone here?

In short... "I reject your reality, and subsitute my own". Somthin' ain't right here.
 
hazydave said:
Yeah, I saw that... I'll admit, I'm suspicious. Particularly because their data, and particularly image capture, also shows the XF100 better than the XF300. That can't actually be possible... three of the same senors, versus the one.

I'm not shocked that overall color would look better than many 3-chippers given enough light, based on state-of-the-art CMOS and 4:2:2 encoding.... my experience with the HVR-A1 suggests that the higher color resolution of a good CMOS chip can help, and that the Bayer interpolation does really affect color perception (you have the same amount of color samples prior to MPEG or AVC encoding), just the accuracy (you will have interpolation errors on some material). I'd like a little more information on this model, too... could be their image processing can account for boundaries in ways impossible before this.

But why the 12lux XF100 looks better than the 12lux XF300... something ain't right.
Click to expand...

It looks to me like the 12lux image for the XF100 has far more noise reduction than the XF300 - lots of smoothed over chunks with plenty of artifacts on the edges. I see that the XF300 looks warmer than the 100, but I can't speak for the color balance, I'm not sure what the settings were. It's not an issue in real-world shooting. Otherwise, they look pretty comparable - I don't see a clear winner. However, in other areas tested (luma resolution, chroma resolution, moire & on-cam mic quality), the XF300 handily tops the XF100.

Despite how the 12lux image looks (and I'm not sure of the settings for that image) here's how the low-light specs compare on the two:
XF100: Full AUTO mode: 4.5 lux (Shutter speed 1/60, Gain +24dB); Manual mode: 1.6 lux (shutter speed 1/4, Gain +33dB)
XF300: Full AUTO mode: 4.5 lux (Shutter speed 1/60, Gain +21dB); Manual mode: 08 lux (Shutter speed 1/4, Gain +33dB)

The only thing that I don't like about the slashcam tests is that they're done in interlaced mode. When the XF300 is shot in interlaced mode the the line-averaging process used for deriving the interlaced fields results in a loss of detail. Canon rates the xf300/305 at over 1000 TVI, and the XF100 at 900 TVI, so I think it's safe to say that the 3-chip 300/305 will still outresolve the 100/105.

All in all, I don't see anything terribly surprising in the comparison. The XF100 is a great camera, and the XF300 is a great camera. Their performance as shown on slashcam is indeed comparable. However, with the 300/305, you're also paying for a lot of features not quantifiable from the slashcam tests. For example, the 3xx has a faster, longer & (more than likely) better performing lens. It has a larger, higher resolution LCD & viewfinder than the 1xx. The 3xx can be upgraded to PAL, at cost. The 1xx has only 1 lens ring, no repeatable manual focus with hard stops, and the built-in ND filter is controlled by camera - auto or off only. So, yeah, slashcam definitely does not tell the whole story, and this should be kept in mind when comparing any cameras. You always get what you pay for, but you should pay for the features and quality you need.

I don't think the slashcam tests should be taken as god's word, but they are a good indication of what performance will be like, and worth paying attention to. I would never recommend purchasing a camera before you get the chance to see real-world performance, though, and ideally handle one for yourself.

All that being said - I apologize for hijacking this thread, which should be about the XA10.
 
Last edited:
Look carefully at the Slashcam results. The XF100 seems to have more chroma noise than the XF300, and more of a watercolor look due to NR, and the color saturation is also lacking, as well as definitely less red channel. Not a better image at all.
 
I honestly think that this seems to be a neato camera for certain types of work. The sound controls are really neat, in camera, it has all sorts of directional modes, and then with the handle, it seems to have all the hook-ups you need for a bunch of mics, and in all honesty, if there isn't some sort of AGC thing going on, this would be a good set-up for people who truly want a all-in-one package. Then, there's so many millions of things that really justify this cameras price tag; night vision, all sorts of creative modes, focus ring, great screen technologies (zebra, focus pulls, ect ect), all sorts of manual controls like aperture and good bokeh, and to top it off, the camera is tiny as crap. I'd would seriously vote this as a camera that may revolutionize the portability of doing something like a Michael Moore documentary - this camera will make things more appealing to more people who just want to get out and shoot, with great sound and footage all in one, and for $2,000, it is jam packed with billions of useful and needed things. This is exactly where video technology should be heading.

One negative though really is that they need to get the ball rolling with depth of field options, if a camera like this doesn't have useable depth of field at all, it really puts a negative twist there for a camera that's $2,000. Really though, this camera is still one that I feel could replace so many other cameras in the right hands, it could still be used as a newscamera, used for certain types of documentaries, could be a perfect student camera for school systems to invest in (for students who need to learn about sound and picture all in one), and also, really the size of it is downright trendy and for the first time, the actual features of a camera are more tremendously overwhelming than the size. You could pick this thing up and practically stuff it in a large coat pocket, yet it'd take you a darn good long time to really get used to the features. I like that, so very much.

All in all, I wish it'd have some good ol' depth of field, but if not, I'd say this type of camera would really be a great package for certain types of work. Documentary, reporting, student and faculty recording, certain types of commercials and product ads, ect etc. I mean, you walk into an old high school and they some hunky tape camera sitting around collecting dust, at least this is a good alternative they can be updating their video departments with. After all, school systems need to get on the ball in the filmmaking department, you walk into a school and there's literally over 500 $1,000 computers for every teacher and student, yet there's not one camera to use in a filmmaking class. It sucks. Maybe as technology gets more "all-in-one" or "trendy and actually useable", schools won't be so afraid to invest in something that will really appeal to filmmaking students.

Really, they should be investing in 60D's too, isn't that hard.
 
Low light compared to DVX100b?

Low light compared to DVX100b?

Hey all,

Haven't been here in a while but I'm still rocking my DVX100b cams for dance recitals, etc.

Just wondering, with the larger sensor, how this model and it's slightly less expensive brother compare to DVX100b...just in lighting situations.

I'm talking about a decently lit stage or venue.

I'd love to stay Panny but these Canons seem a reasonable alternative for me to go HD with.

Thanks for listening.

ian
 
hazydave said:
I pretty much have decided that Slashcam is a work of fiction... who are these guys, and why does anyone trust their results.
Click to expand...
Something else I noticed that makes me too very suspicious - Take a look at the distortion when you have the XH A1 selected. It just isn't there like it would be if it were real life. I'm assuming the whole point of that image is to show the barrel distortion at full wide, but it's inconceivable that the A1 is at full wide in that image. Then, when you change it to the XH-A1S, the distortion shows up the way it should. I don't remember Canon changing the lens for the S model.
Of course this is just one example but like others have said, just more evidence that this should probably be taken with skepticism.
 
Moseph said:
Something else I noticed that makes me too very suspicious - Take a look at the distortion when you have the XH A1 selected. It just isn't there like it would be if it were real life. I'm assuming the whole point of that image is to show the barrel distortion at full wide, but it's inconceivable that the A1 is at full wide in that image. Then, when you change it to the XH-A1S, the distortion shows up the way it should. I don't remember Canon changing the lens for the S model.
Of course this is just one example but like others have said, just more evidence that this should probably be taken with skepticism.
Click to expand...

Camcorderinfo's review of the HF G10 also has it as better than the XH A1s, so Slashcam might be right,
 
I have an XH-A1 and an HF-S10 and HF-S21 that I show sample clips from. To the untrained consumer....most customers can't tell me from what camera the footage came from??? And as far as I know...they did not make lens changes from the A1 to the A1s.

Just a side note...I got a note from Canon a couple of days ago, After 10 years...they are finally retiring the GL-2! I can't believe it lasted as long as it did! 10 years! Isuspect that the A1s and G1s are soon to follow. The tape drive companies have just about given up on making the drives. I hope everyone likes SDXC, CF or SSD!
 
Take a look at the Camcordinfo review. Compare the test setup shots of the HF G10, which is simply a A10 without the XLR handle, to the shots of the other cams, including the DSLRs. It appears to be better than all but the 5d, and it may be sharper than the 5d with different chroma.
 
You must log in or register to reply here.
Back
Top