Canon XL-H1 + NanoFlash or Canon XF305?

Mikungfu

Mikungfu

Active member
Hi Everyone,

This is a long post so blease bear with me and thank you for your patience in reading through:

I have been eagerly waiting for Panasonic AG-Af100, because having a DSLR Shalow DOF and Low-Light capabilities, in a Pro Camcorder form factor with all the video functionalities, wissles and bells, has been my dream for quite sometime.

I have always hated all the rigging, aliasing, poor audio functionalities, etc... Of the DSLR solutions.

I work in narrative indie film making, publicity, documentary, events and institutional video.

The only problem I invisioned with a AG-AF100 is that I'm almost sure it will be better for narrative video production or controlled environment shooting, but that great for other more dynamic and unpredictable situations, where the shalow DOF can be more of a con than a pro.

I am ready and willing to invest on a 2 camera set. I used to work with a pair of Canon XL-H1 for several years and loved it. I recorded on tape, though...
Unfortunately an unexpected problem in my life caused me to sell them both about a year ago... And now I'm ready to rebuild my camera set.

These were the possibilities I had in mind:

Possibilty #1: A pair of Panasonic AG-AF100 + a pair of NanoFlash.

* Pros: Shallow DOF with all the wissles and bells of a Pro Camcorder and its form factor.
* Cons: Having to wait till till probably February 2010 and probably unsuitable for events and other less controlled shooting.

Possibilty #2: A pair of Canon XL-H1 (used) + a pair of NanoFlash + SGBlade 35mm Adaptor.

* Pros: Possibility of selling them for the price I bought them, once the AG-AF100 comes out. And then buy the AG-AF100 (if I still have the money:undecided). Possibility of real time backup on tape when shooting. Possibilty of having both shallow DOF for narrative and deeper DOF for events and stuff. Also... I love it's form factor.
* Cons: Bad low-light behaviour CCDs. Although it has natively interchageable lens, their almost useless... There is only a wide angle lens for this camcorder, and the DSLR lenses are not that compatible (crop factor, vigneting, etc...). For that, I would need to buy the SGBlade and that brings to issues: Even worse low light performance and once I buy the pannys, these adaptors would be useless.

Possibilty #3: A pair of Canon XF305 + SGBlade 35mm Adaptor.

* Pros: Possibility of not needing nanoflash so soon, since it's 50mbps is enough for broadcast, chromakey, etc... Possibilty of having both shallow DOF for narrative and deeper DOF for events and stuff. A great lens and a great codec. Much better low-light performance than the XL-H1 and it's advantage for the use of the SGBlade.
* Cons: Fixed lens (overcome by the adaptor). Expensive as hell. I prefer the semi-shoulder mount in the XL-H1's form factor. Rolling shutter and skew issues.

SUMMARY: I want to choose a new pair of cameras. I love shallow DOF but cannot be stuck with it, since I don't do just narrative stuff. I need the possibilty of having both shallow and deep DOF for different situations.

The AG-AF100 is yet semi-vaporware and we're yet to see how it turns out. Will it have deeper DOF with some lenses? Will the auto-focus of the Lumix Micro 4/3" lense have a good effective auto-focus? will it be suitable for events and shoot and run handheld shooting? Also the AVCHD 24mbps is not as good as the XF codec 50mbps.

The XF305 seems to have outstanding image quality, great low-light work and powerful functionalities. But the fixed lens have always bothered me and I don't love it's form factor when compared to the XL-H1. It's native codec is enough for broadcast with no need for NanoFlash. It's CMOS should perform much better with the SGBlade than the XL-H1 CCDs.

The XL-H1 have a beautiful form factor and can be bought and sold without loosing money. Although it's semi-useless, it does have interchangeable lens. With a NanoFlash we can capture higher bitrate than with the XF305... But I believe the color space is still inferior in the XL-H1, even with the NanoFlash, right? The lens is also inferior to the XF, in terms of Chromatic aberration and distortions.

So, after this long post... Can anyone give more input, info, opinions and feedback that may help me reflect on these 3 possibilities? I am definitely going to make the purchase decision in the next days and, all this brainstorming has been killing me. I really some light shed on this subject to clarify my ideas.

Thanks.
 
take a look at these videos

very nice looking dof
http://www.vimeo.com/13780183

steadicam
http://www.vimeo.com/14610176

low light footage
http://www.vimeo.com/14749439

some miscellaneous footage
http://www.vimeo.com/14723949

and more dof shots
http://www.vimeo.com/13282137

the only reason i chose not to get the xf300/5 myself was i felt i would need a little bit more power in terms of low light

also the xf 305 has a 4" lcd vs a 2.4" lcd

and at b&h its 500 off right now
http://www.bhphotovideo.com/c/product/684198-REG/Canon_4454B001.html#specifications
 
Thanks for the videos, wannabapilot! But how would you compare the SDI output of a Canon XL-H1 recording to Nanoflash, with the image recorded by the XF305/XF300 in compact flash cards (XF codec Mpeg2)?

Do you think XF305/300 still has better image in CF than the XL-H1, using SDI out?

...And with the XL-H1 we could be either talking about it using the 20x bundled zoom lens or the 6x wide angle lens....
 
The XF300/305 has a superior picture to the XLH1 even given SDI output recording. The lens is better than the kit XLH1 lens (i've not tested against the wide angle), the sensors are far cleaner, they are full raster and resolve over 1000 TVI lines of resolution (vs the 850-ish of the XLH1).

Remember, even with an external recorder, any camera is only as strong as it's weakest link. If the recording codec was the only thing separating the XLH1 from the XF series, then the story would be different. But there is much more that sets the XF line well above Canon's HDV line in terms of overall IQ.
 
Barry Green, HVX-HPX extraordinaire, reviewed the AF100 and loved it, even going so far as to say it's the death-knell of HD-DSLRs. Definitely worth looking at. But if you want 4:2:2, you're gonna need a Nano or some other external recording device.

The XF300/305 line seem to be outstanding cameras. BBC approved, and that's saying a lot. You won't need a Nanoflash with these. Save you some $$. You can even go for the cheaper XF300s, unless you need to sync timecode.

XLs are HDV - nope. Even with Nanos - nope.

Tough decision. If you can wait til later this year, you can check on the feedback from private owners of the AF100, see some video examples.
As it is right now, if I were in your shoes and I needed cameras quick, it'd be the XF305s, no question.
 
The XL-H1 has a 1440 wide interlaced sensor, which is fine for HDV (also 1440 wide.)

However, compared to full raster its 25% less picture data and half the color data at half the data rate.

Recording to the Nanoflash does not create new data. Any full raster camera like the EX1, XF30X, etc will have the potential to create a much clearer image. The Nano on an EX1 helps certain shooting situations. The Nano on an XLH1, not so much.
 
A lot of people are thinking of pairing a three sensor XF300/305 with a single sensor XF100/105 as the B cam, which is what Canon said it was designed to do.. BUT no sample of image quality or low light suitability. Both full raster.
 
Last edited:
Duke M. said:
Recording to the Nanoflash does not create new data. Any full raster camera like the EX1, XF30X, etc will have the potential to create a much clearer image. The Nano on an EX1 helps certain shooting situations. The Nano on an XLH1, not so much.
Click to expand...
I always thought the SDI out bypasses the HDV compression, why would Canon make it so that the data gets reduced to a fraction of what it is coming straight from the CCDs and then make that tiny amount of information uncompressed?
 
Moseph said:
I always thought the SDI out bypasses the HDV compression, why would Canon make it so that the data gets reduced to a fraction of what it is coming straight from the CCDs and then make that tiny amount of information uncompressed?
Click to expand...

But what if it were not a full raster cam and the signal from the CCD was 1440? See Duke M's reply in this thread.
 
Also the XH-A1 is only capable of resolving 1440x1080 in 60i, putting it in progressive mode drops the resolution farther, so you would be much better off with the XF300, which has a pretty good codec.
 
philiplipetz said:
But what if it were not a full raster cam and the signal from the CCD was 1440? See Duke M's reply in this thread.
Click to expand...
Yes, I'm aware of the fact that it won't make up any new pixels, but the nanoflash should be able to record much more overall data, getting rid of the compression artifacts of HDV. I don't disagree that the XF would be vastly superior but a tapeless workflow isn't the ONLY thing you get out of pairing a nanoflash with an XL H1, unless I'm missing something.
cgold said:
Also the XH-A1 is only capable of resolving 1440x1080 in 60i, putting it in progressive mode drops the resolution farther, so you would be much better off with the XF300, which has a pretty good codec.
Click to expand...
It's the XL H1 in discussion here, not the A1, but on that subject I actually remember reading that someone tested the A1 and in 60i it could resolve 1015 lines, in the F modes it was around 950. Not bad for a pretty good looking "fake" progressive coming from interlaced CCDs.
 
Sorry so used to saying A1, but the sensors on the two are identical, the A1 just lacks the SDI out.
 
Moseph said:
It's the XL H1 in discussion here, not the A1, but on that subject I actually remember reading that someone tested the A1 and in 60i it could resolve 1015 lines, in the F modes it was around 950. Not bad for a pretty good looking "fake" progressive coming from interlaced CCDs.
Click to expand...

Unfortunately, I think this is a bit optimistic. The Texas HD Shootout from 2006 showed that the XL-H1 (and with the same sensors & processing, more than likely the XH-A1, too) can only resolve 800x700+ lines of resolution in 60i, while in frame mode taking a hit down to 800x540.

http://www.dv.com/article/17652
 
Well I'm pretty sure we have a res chart in the production office where I'm going to school, I'll see if I can get down there and test it out in the next couple weeks or so and report back.
 
there are things a CCD camera can do a CMOS camera cannot, and vice versa. i would argue a different option. XF305, SG-Blade, and an XL-H1/XH-G1 to cover all bases. I use both CCD and CMOS cameras, 35mm adapters (Letus Extremes), and both systems are invaluable in different ways.
 
Moseph said:
It's the XL H1 in discussion here, not the A1, but on that subject I actually remember reading that someone tested the A1 and in 60i it could resolve 1015 lines, in the F modes it was around 950. Not bad for a pretty good looking "fake" progressive coming from interlaced CCDs.
Click to expand...

I'm afraid that 1015 line pairs is far above the Nyquist limits for A1 or H1 sensors. (720 under perfect conditions and nothing is perfect.) Someone had the camera too close to the chart in order to get those kind of numbers. Its just physics. You need a full raster 1920 sensor and file format to get over 800. Then full raster on a 3 chip camera to get up to ~1000 (960).
 
Last edited:
Duke M. said:
I'm afraid that 1015 line pairs is far above the Nyquist limits for A1 or H1 sensors. (720 under perfect conditions and nothing is perfect.) Someone had the camera too close to the chart in order to get those kind of numbers. Its just physics. You need a full raster 1920 sensor and file format to get over 800. Then full raster on a 3 chip camera to get up to ~1000 (960).
Click to expand...
I figured those numbers did seem a little high, but out of curiosity, how does the number of vertical lines affect how many horizontal lines that you can get? You say you need a full raster sensor, now obviously something like the HVX or HMC150 wouldn't be able to put out over 800 since the CCDs are around 960x540 (I think), but the A1 does have 1080 horizontal lines on the CCDs. If the lens was really perfect, or as close to it as physically possible, couldn't it get close to there? Or is the main thing taking a toll on the resolution the HDV compression?
 
Well as someone who uses a Canon XL H1S/Nanoflash combination everyday, I can tell you there are massive benefits in image quality over just HDV from the camera. The resolution is indeed increased... I think it has to do with the fact that there is horizontal pixel shift happening in the camera that is only capable of being recorded through the HD-SDI port (and as long as you record to a full raster codec like what is used in the Nanoflash). Just like how some Panasonic cams get extra resolution from their sensors. But in this case, the Canon starts with much more res and only has to fill in the horizontal direction. Since the natice res of the sensors fills the HDV spec 1:1, this extra res can only be obtains by recording out of the HD-SDI port, to something like the Nano.

Plus when youa re shooting in the I-frame 200mb+ range, the images really pop out. It is a great combination that shouldn't be overlooked. I actually was at this crossroad: XL/Nano or XF305? After shooting with both, I decided to stick with the XL/Nano combination as overall I felt it was better for me. Not that much better image quality from the XF cams IMO.
 
Last edited:
Moseph said:
Yes, I'm aware of the fact that it won't make up any new pixels, but the nanoflash should be able to record much more overall data, getting rid of the compression artifacts of HDV. I don't disagree that the XF would be vastly superior but a tapeless workflow isn't the ONLY thing you get out of pairing a nanoflash with an XL H1, unless I'm missing something.

It's the XL H1 in discussion here, not the A1, but on that subject I actually remember reading that someone tested the A1 and in 60i it could resolve 1015 lines, in the F modes it was around 950. Not bad for a pretty good looking "fake" progressive coming from interlaced CCDs.
Click to expand...

Yes, I read this too about the 1015 lines. It was from the Alan Roberts BBC test of the XHG1. He measured from the HD-SDI port and not HDV. I believe you can find the report on the BBC site somewhere still.
 
You must log in or register to reply here.
Back
Top