Here's what is in the box:
- Thread starter dashwood
- Start date
stephenlnoe
Well-known member
Barry_Green
Moderator
Incorrect -- the split-screen thing is most definitely on the recorded images.mezelf27 said:
Barry_Green
Moderator
The first I saw of the split-screen mode was from a "PAL" camera. I've also seen the effect on the "NTSC" camera, so it definitely affects all of them.
I haven't had a chance to test the lens at long focal lengths. I also haven't had a chance to confirm if the split-screen effect (which I only witnessed once) transfers into the m2t over firewire, or is just on the analog outs.Rocketeer said:
I believe the first limited shipments were sent out to all the Canadian dealers with pre-orders last Friday. I know that this whole first shipment was QC'd by an engineer from Japan, and that was the cause for the initial delay here. I was told that a second shipment is being prepared now. (this is all second hand information from my dealer - I have no inside information)
I assumed the U.S. shipments had gone out last week because there seem to be alot of people posting footage all over the various newsgroups.
Maybe these are insiders with pre-production models, or are actually the European model.
I feel really guilty that I picked mine up on Friday, (it is now Wednesday) and I haven't had a chance to shoot more than 4 minutes of footage. I know there are a bunch of you out there who can't wait to get your hands on it.
Strangways has already volunteered to help me shoot some tests, as well as side by side against XL2 and DVX100A as soon as I have a chance.
If anyone else wants to lend a hand and you live in the GTA (we are specifically shooting tests in Hamilton/Burlington Bay) please let me know.
I may have some time Sunday or Monday, but we'll see.
Tim
J.R. Hudson
Spirit Animal
Any footage yet?
How long does it take to film something? Come on Tim!
How long does it take to film something? Come on Tim!
John_Hudson said:
I had it in the back yard for 5 minutes yesterday while my son was playing in his sandbox. I wouldn't consider it "test footage" but I could upload a bit to my server in a couple of hours. I tried various gamma settings, as well as frame rate settings.
Tim
Nate Weaver
Veteran
iDisk!
The real problem is how long it takes to post and upload something!John_Hudson said:
dashwood said:
OK John.
Here it is @ 1280x720 in H.264 (80Mb)
You’ll probably need a G5 and QT7 to play this at 24fps.:
http://homepage.mac.com/timdashwood/.Movies/HD100clips.mov
Here it is @ SD anamorphic 480p24 MPEG2 (128Mb for the slower processor people, or maybe you can burn this to a DVD
Update Sept 10 2005: I have removed the m2v because I need the space to upload some more m2t
Be sure to RIGHT-CLICK (or CTRL-click on MACs) these links and save them to your computer. They are BIG!
I should preface this by saying that my primary excitement about the HD100 is the ability to shoot slo-mo at 50 and 60 fps. I don't care that it is in 720x480 (or 576 @50) because I will mostly be using it for music videos and a direct-to-DVD action film I'm shooting in October.
That is why the majority of what you will see is shot in 480p60. I wanted to make sure my theory would truly work. I've done it many times with the DVX and 30p - slow-mo to 24P, but 60 will be awesome! Now I don't need to rent a Varicam.
Here's my current workflow: I'm using Mac OS X 10.3 Panther, with Quicktime 7, but I do not have the FCP5 upgrade yet.
In the case of the slo-mo clips I shot at 480p60, I captured them with HDVxDV (free trial version) and then used MPEG streamclip to export them to 960x720 60P with Apple's DVCPRO HD 720P codec.
I then opened each clip in Cinema Tools and "conformed" them from 59.94 fps to 23.98. Hence real in-camera "slow motion" shot at 60 frames per second. This is an instant process in Cinema Tools - it just tells the quicktime file that its frame rate is now 23.98.
This was by no means a scientific test, but I do remember I had the Cinema curves and matrix turned on and I set my zebra to 100% and then exposed my highlights using that as a guide.
The 720p24 clip was captured with HDVxDV and converted to DVCPRO HD quicktime codec in the native resolution. After I edited the clips I wanted in FCP4.5, I recompressed into H.264 codec at 1280x720p24. (I created the 480p MPEG2 from the original 480p m2t files.)
Here's the description of each clip:
00:00
A clip of my 9 month old captured in 480p60 (with a 1/100 shutter I think) uprezzed and slowed down to 720p24. I was using "macro" mode. I put this here to show the whip pan slowed to 24.
00:11
Also captured 480p60 and uprezzed/conformed to 24, but I colour corrected this because it was very underexposed when in the shade. I really cranked the mids and added saturation, but you can really see compression artifacts. I'm not sure if this is indicitave of HDV compression in lower levels, or if I introduced it by uprezzing in streamclip? I also stuck it on the end without colour correction so you could see the original.
00:32
Here's a clip of my other son captured in 480p60 uprezzed (with QTpro resizing) and conformed to 720p24. I was telling him to run around, but he didn't understand why. Notice the motion of his hair blowing in the breeze near the end. I was adjusting the iris during this shot until the hotspot on his hair was reading 100IRE according to the zebras. White clip was set to 108%.
Take note of how the MPEG compression "trails" when he swings his arms up and down. Hopefully this is just a result of current HDV capture software not translating 6-frame GOP properly.
01:16
Playing with sand and water, captured in 720p24. The problem here is that the file is captured at 59.94 and I can't seem to remove the advanced pulldown because version 4.5 of FCP is all based on 29.97 interlaced.
However, by dropping the 59.94 clip into a 23.98 timeline, the pulldown seems to be removed automatically (requires a render.)
Can anyone else figure this out, or should I just hold my breath for an FCP5 HD100 update? The HD100 captures 1280x720, but this clip has been downsized into 960x720 DVCPROHD and then output back at 1280. I also think some DVCPROHD artifacts were introduced when it was rendered into the 23.98 timeline. The motion blur looks a little weird to me. The shutter may have defaulted to 1/24 when I put it back in 720p24 mode (stupid default!) I'll have to make a little note on the camera to always remember to set the shutter to 1/48!
Here's the link to the original HDV 720p24 m2t mpeg transport stream of this clip. Make sure to Right-click this clip and then save to your computer. You can then view it in a Mpeg transport stream viewer. It is about 48Mb.
http://homepage.mac.com/timdashwood/.Movies/alex720p24.m2t
01:45
I also shot some stuff in NTSC DV 24PA, and captured with FCP4.5, so I could compare with DVX100 and XL2. I have to say that I am very disappointed in the way the camera processes DV. It seems to leave MPEG ghost artifacts on fast pans. You can see it on the yellow plastic when I whip left or right. I've never seen this with the DV codec. Shown here in native 720x480 stretched to 16x9 within 960x720.
02:25
Original SDp60 uprezzed clip2 without colour correction.
That's it for now. I wanted to wait until I had something better, but oh well.
Tim
mezelf27 said:Nice made foam. Did you cut that yourself? And how in hell's name did you accomplish that? - It's not that I'm not handy, but these things always get messed up by me...
You can buy Pelican cases with 1cm square perforations, and then it is easy to pull out the bits you need to make anything fit.
That's my current observation, but Barry disagrees. I've posted some footage, but it was all shot with the gain off.
Tim
Nate Weaver
Veteran
dashwood said:
Hang out for an update. I've talked with David Newman of Cineform on another board, and he's explained what's happening in the 24p streams and what it will take for NLEs to work with it...it's not a big job for them to be updated.
Barry_Green
Moderator
I've seen it in recorded footage. Try just capping the lens and recording some 18db gain stuff, and play it back -- you'll see that it definitely affects the recorded footage.dashwood said:
There's another post on DVDoctor that shows the split-screen effect in captured footage as well. So, don't get your hopes up -- the split-screen most definitely affects the recorded image.
Downloading the clips now...
Barry_Green said:
Hi Barry,
I did a split-screen test at +18db and I can definitely make out a level difference of the firewire captured footage split left/right. I also found a white pixel!
However, it is not nearly as pronounced as on the NTSC live signal.
I've uploaded the m2t to my server. 38Mb. Download it here:
http://homepage.mac.com/timdashwood/.Movies/split-screen-test+18.m2t
The shot starts with the lens cap on, but you can really see it best when I zoom to the blue wall.
Honestly though, I very seldom shoot with any gain for most of my applications. It comes in handy sometimes when I'm shooting documentaries, but I don't remember ever shooting anything for broadcast at more than +6db (even on my betacam.)
So, this probably won't affect me now for my current projects, and it will be great if they clear it up in a firmware patch.
Tim