View Full Version : XLH1 vs F900 (NAB test frames)
Adam Burtle April 27th, 2007, 01:07 AM NAB has passed now, and for those who weren't able to attend, we were showing several scenes of sample footage (http://www.colorspaceinc.com/hexbooth2.jpg) in our booth (previous photo taken in a rare, pre-9am quiet moment). The footage was all shot in the first week of April (which seems like a decade ago now!). XLH1 head to head with F900, Viper head to head with 35mm Kodak Vision2, etc.
Now that we've had a few days to decompress, I figured I'd post something for those H1 users who didn't have a chance to attend. I just went into my afterFX comp and grabbed a couple of samples, but this should give you an idea of how your H1 will do in place of an (~$75,000) F900 (I think the lens on the 900 alone cost more than 2x what the XLH1 does). The short answer is that both cameras look quite lovely once you take tape compression out of the mix (recording 4:2:2 over HD-SDI). My own personal opinion is that the H1 is marginally noisier, but during play, its noise actually has a nice random/organic quality.. filmgrain-like.
Everyone on our crew, including the DP (and even myself) were quite suprised at how close the H1 is to the 900 out of the box. The only major tell is the depth of field.. and we're eagerly awaiting the HD mini35 adapter from P+S, as that adapter may prove a nice solution for those who want wide angle pov and shallow dof, yet don't have tons to spend on a 2/3" format camera.
Capture #1 (http://www.colorspaceinc.com/xlh1_1.jpg)
Capture #2 (http://www.colorspaceinc.com/xlh1_2.jpg)
Shameless plug: "we" is Colorspace, Inc... (http://www.colorspaceinc.com) we've announced two almost identical field recorders, the ICON and the INDI. ICON is uncompressed, and meant for Viper, etc. The INDI is high bitrate compressed recording only, and is targeted at the XLH1/HD250 market. Retail availability will come later in 2007, and target pricing for the INDI is currently a bit over $6k + media packs will run around $3k.
John Benton April 27th, 2007, 11:37 AM Adam,
Thanks for posting,
What is the side by side comparison in the second Photo
(the Left image being a lot nicer than the Right)
Thanks,
J
Adam Burtle April 27th, 2007, 01:26 PM What makes the left image nicer? Apart from depth of field? (I'm curious as to your take-- not disputing one way or the other)
The #2 photo is essentially a 250% enlargement of an area from the frames scene in photo #1.
John Benton April 27th, 2007, 01:30 PM the color and more light, less grain in the image on the Left.
I take it that is the 900
whereas the Right side is the H1?
Herman Van Deventer April 27th, 2007, 01:34 PM XLH1 vs F900
Adam,
Thank you for the comparison.
Mind boggling ! I really can't see how the difference justify app. $ 60 000-00.
Herman.
Darrell Essex April 27th, 2007, 01:36 PM So is the H1 on the right?
Darrell
FIRST CINEMA PICTURES
Adam Burtle April 27th, 2007, 01:42 PM John / Darrell -- yes, the H1 is on the right in both. I was trying to get my answer far enough down the page so as not to ruin the surprise for some ;)
Pekka Uotila April 27th, 2007, 01:56 PM Very nice Adam,
I guess I was able to guess the left and right from your hint about the dof. From the detail picture (on the right) it is possible to see some more CA around the handles below the man's head. Yes, with money it is possible to get more quality and better lenses but...
XL H1 seems to have little bit more of red/magenta on the skin.
It would be nice to hear what lenses and f stops you were using and what type of settings you were using XL H1.
Some more pictures would be very interesting if it is possible.
And yes, your Indie Icon thing will be something very interesting for me I believe.
Thanks
Adam Burtle April 27th, 2007, 02:28 PM Pekka.. in the interest of a "fair" comparison we tried to use both cameras just as they came. The XLH1 had the gamma extension turned on, but otherwise nominal settings. I believe the color differences could be tweaked in camera pretty easily (if one were trying to shoot with both and intercut). Shooting with just one, you could leave the color however and just correct in post as normal.
I don't have all the shot notes handy, so I can't tell you the exact lens on the 900, but it was a Canon zoom. H1 had the 20x on it (except for one scene, we shot one angle with the 3x Standard Def lens-- we were showing this at NAB, and the lens performed beautifully.. test charts comparing it to the 20x show they are VERY close).
F900 was at F4 1/3, H1 was at F4, both 0db gain, 180deg shutter.
I'll post a second thread later today with some more stills when I have a free moment to extract them.
Daniel Patton April 27th, 2007, 07:50 PM Adam, Great stuff, thanks for sharing.
The F900 should damn well look better at 250%, considering the price tag! ;) I can't help but wonder if the image detail on the H1 was turned down a little might the noise be reduced some and the images a bit closer. So you are saying that you have done no CC at all, and this was roughly straight out of the box for both cameras? If so then I'm shocked at how close the overall color is between the two. Other than dof I also agree that these could be cut together with no major concerns.
Adam Burtle April 27th, 2007, 11:09 PM Daniel.. the footage was cut, composited, and finished completely inside a default install of After Effects 7. Speedgrade (etc) never touched it. It did have some very minor color tweaks, but we're talking like "+10% saturation" .. stuff like that.. just to correct for minor differences that could have been addressed in-camera. Also, the "CSI" scene (what these stills are from) mixed HMI, incandescent, daylight, and fluorescent light* hahaha.. so using only a dab of CC is pretty phenomenal (and I credit our DP and crew for just a great job with mixing light sources).
*disclaimer: this is from memory, I am not a DP, and I don't want to speak for Bryan-- so he might show up here and say "adam is an idiot, there was no HMI in that scene.." or whatever.
also, while we're talking about the minimalist approach, I may have forgotten to mention that besides being posted entirely in AFX7, we rendered to CineForm (high quality mode) and played the footage realtime on Mac Minis. No fancy raid arrays, just $900 mac minis playing beautiful 1080p footage in realtime. I wanted to do this specifically as a challenge to prove that you don't need to break the budget to work 1080p. 4k is very exciting, but still very expensive. 1080p is a reality for even low budget work at this moment. Now, 1080p CineForm is right at the limit of what a Mac Mini can handle, but it's a nice proof of concept.
Gary McClurg April 28th, 2007, 06:24 AM Adam...
A few questions...
I'll go through my NAB stuff... I know I saw your booth... but nothing on the Canon... but who knows might have been the day my feet and back were tired... which was almost every day...
Okay...
How much is the rental?
How much storage space will be needed... for example I like the easy math approach... 1 gig equals a minute... five gigs equals 1 minute...
And last but not least could we rent in a month or two...
Thanks...
Mouayed Zabtia April 28th, 2007, 12:01 PM Waaaaaw so big different in DOF so now i get understand the different between 1/3 and 2/3, can i ask if sony 900 have that nice DOF or all 2/3 camcorder have it?
Brian Drysdale April 28th, 2007, 02:26 PM Of course, the F900 isn't at its best straight out of the box, there are gamma curves and menu adjustments that improve it considerably. The weakest part of the F900 is the HDCAM recorder on the back end. Same goes for the HDV recorders on the XL H1 and the JVC HD 250.
Shooting with the lenses wide open usually shows how good they are. Commonly, a stop of f2.8 or less is used on a F900 mounted zoom lenses to give a shallower DOF.
Nick Hiltgen April 28th, 2007, 11:06 PM He He He
Adam you've opened a can of warms comparing the Canon to F-900. I'll go ahead and tell you from my experience ahem... http://www.dvinfo.net/conf/showthread.php?t=56347&highlight=xl-h1+f900
For some reason no one likes comparisons. Unfortunately (for everyone else) it seems that you'rs was pretty well done. Most importantly I think that it illistrates the point and the advantage of capturing uncompressed. I think something that might work even better is showing the HDV compression against the uncompressed stills. The next thing would then be to show how quickly the HDV codec falls apart when working in color correction, or compositing.
So what's the dvinfo discount for the INDI?... ;)
Floris van Eck April 29th, 2007, 04:38 AM The INDI looks very nice. However, I think it will place a very high demand on my storage. How much is 100% uncompressed HD-SDI? Like 1,5GB per minute or something in that line? That would mean 150GB for 60 minutes of video.
The problem is that as long as there is no clear winner in the BluRay / HD-DVD format war, and prices do not come down quickly... all the true HD work is really expensive and only affordable if you are really earning your money with it. For me, as a hobby and beginning enterprise, it's too expensive at this moment.
But in the future... this is where we go. Get the RAW signal from the camera and choose your favorite compression scheme to record the footage in.
I would say keep us informed on pricing, availability and details.
Can you replace the harddisk on your own or do you have to order all media from ColorSpace?
Aaron Burtle April 29th, 2007, 11:00 PM Floris,
The uncompressed signal off the XLH1 is 1.485 Gigabits per second. The INDI Recorder takes that 1.485 Gbps signal, and compresses it using MPEG4 compression to around 200 Megabits per second. If left in that compression format you would need around 90 Gigabytes per hour of footage. The media packs which slot onto the INDI do indeed need to be purchased from Colorspace. The media packs are made specifically to dock and communicate with the FPGA recorder. However, you can transfer your data from those media packs to any number of storage mediums you wish to use.
Pekka Uotila April 30th, 2007, 12:38 AM Adam and Aaron,
thank you for all the info sofar,
it is nice (as an XL H1 owner) to see the comparison the way you did it even thought both of these cameras can be tweaked etc. as it is mentioned in this thread already. Personally I have no plans to mix these two cameras.
At the moment I have been matching XL H1 and XH G1, ...it is a bit tough but I am getting promising results... , as a tool for this I have "only" an analog SD Hamlet 302WVR and a good 14" Sony also with Component SDI input. Are you able to say something about how reliable SD scopes are compared to HD scopes? The shooting with these two cams will be in HDV.
Since my use with XL H1 has been mainly in SD reality so far, it would be very interesting to see with my own eyes what means "horribly restrictive HDV compression" , like you say in your site.
So it would be very interesting to see comparison between HDV codec and your codecs. I guess you were recording also on tape with XL H1.
Btw, a friend of mine was visiting at your booth , you had some dynamic range info there. It would be nice to hear if you find something new about this matter in your tests?
Finally, hopefully the "dvinfo discount" will be there one day.
Cheers,
|
|