Kaku Ito's XL H1 video clips -- here's the second batch

[Matt Davis]

Originally Posted by Soroush Shahrokni

I still like the 24P of HD100 more than the 24F of Canon and I still dont know what makes the diffrence but for sure there is something!

I think the 60i stuff looks much sharper than the 24F, and I think this is the work of a deinterlace algorythm!

Nate Weaver and Tim Dashwood have some beautiful 24P and 30P footage that are sharp yet filmic, and the motion isn't particularly strobe-like, whereas everything I've seen from deinterlaced footage has gone ever so slightly soft and stuttery.

I guess the difference is what happens when one designs a Progressive camera from the start, vs the video oriented XLH1 and Z1.

But that XLH1 lens... Phew. Now, how can we bolt that onto the HD100, I wonder? ;)

[Kaku Ito]

Originally Posted by Steven White

Wow!!! That chroma clip is gorgeous.

As for the aberrations, the star shaped flare is a (rather pretty!) property of the lens itself. The rainbows you see around the flares are called "chroma fans" - also a lens property. The vertical lines you see associated with the flares are an artefact associated with how CCDs are sampled.

Chromatic aberration doesn't really occur in this clip until your completely zoomed out. (Aberration occurs primarily at the edge of the lens). Near the end of the video, look at the gaps between the leaves and the sky at the edges of the picture. Here you might see some slight green/magenta shifting at the edges of the leaves. This is the chromatic aberration, and it is really quite minimal. To escape chromatic aberration with this camera, always operate it at slight zooms.

-Steve

Thank you for your great observation.
It seems to be hard not to use the widest angle since the standard lense does not give you wide enough view! But it is good to know how to avoid it. Thank you Steve.

[Chris Hurd]

Originally Posted by Kaku Ito

I'm uploading a 1GB 1080/60i uncompressed file right now, you can cut it to make the download reasonable. It's a Kona 2 file. I think you can download the playback codec form AJA.

This may have to wait just a short while -- I will need some outside help with this. My hosting provider will throw large, hurtful things at me if I make a 1GB clip available for public download. If there's an "Inner Circle" member here who wants to take a crack at editing this file down to 200 or 300MB or so, please contact me off-list by private email. I won't be able to work with it myself on this machine and my editing system is down right now.

[Kaku Ito]

Originally Posted by Chris Hurd

This may have to wait just a short while -- I will need some outside help with this. My hosting provider will throw large, hurtful things at me if I make a 1GB clip available for public download. If there's an "Inner Circle" member here who wants to take a crack at editing this file down to 200 or 300MB or so, please contact me off-list by private email. I won't be able to work with it myself on this machine and my editing system is down right now.

Chris,

If that is the case, it is faster for me to cut it out. I will work on the file tomorrow to make it shorter. If anyone want the clip as soon as possible, they can download it from my site.

[Chris Hurd]

No problem Kaku, I realize 200-300 MB is not much time for uncompressed video but I really am forced to keep the download file sizes manageable. Sorry about that,

[John Jay]

Is this not the same clip as before but longer, with the parrot chirping in the background?

I was hoping for some reflected light tests such as from a checkerboard especially at widest aperture where all the problems are likely to be.

this pinhole aperture F9? shows a lot of smear and also the pattern of the microlens array which is bonded to the CCDs.

would like to see more aggressive testing of the lens, but I guess that will come in time, no doubt.

[Chris Hurd]

Yes, rest assured that formal tests in controlled situations complete with charts, etc. will come later. For now please realize that these particular clips are simply the result of Kaku Ito being very generous in sharing his limited time on the XL H1 with the rest of us. This was never intended to be a formal study. He had the camera on loan for a short while, was able to shoot a few specific things for us and also had his own requirements to take care of. This is not a formal "DV Info Net Case Study" but simply the generous input from one of our members in Japan who was able to put his hands on the H1 for a short period of time. The man has other things to do as well. Don't worry, we'll pursue all the H1 bench-testing you can imagine as the camera becomes available more readily.

[Kaku Ito]

Originally Posted by John Jay

Is this not the same clip as before but longer, with the parrot chirping in the background?

I was hoping for some reflected light tests such as from a checkerboard especially at widest aperture where all the problems are likely to be.

this pinhole aperture F9? shows a lot of smear and also the pattern of the microlens array which is bonded to the CCDs.

would like to see more aggressive testing of the lens, but I guess that will come in time, no doubt.

John Jay,

I'm sorry that I could not work on the aberration test, but it threw me off that the 24 frame portion of the tape could not play with Sony cams, that limited my time to shoot more. I thought that even bikeseq files showed the problems, I thought it seems to have the problem already and not necessary to do the test. I guess it is obviously you wanted to see the poblems in diffeent degrees.

I hope Canon would provide me the cam again for a longer period to do that test for the next time.

[John Jay]

Originally Posted by Kaku Ito

John Jay,

I'm sorry that I could not work on the aberration test, but it threw me off that the 24 frame portion of the tape could not play with Sony cams, that limited my time to shoot more. I thought that even bikeseq files showed the problems, I thought it seems to have the problem already and not necessary to do the test. I guess it is obviously you wanted to see the poblems in diffeent degrees.

I hope Canon would provide me the cam again for a longer period to do that test for the next time.

Hey Kaku

no worries

relax and good night

[Daniel Broadway]

I am a visual effects student currently in college. I like to do lots of blue/greenscreening. I found HDV from the Sony to be unusable as there are waaay to much MPEG artifacts in high motion scenes.

I expected the same from this camera, but upon watching the clip of the guy on the bike, I can find very little artifacts, even when viewing the chroma channels on their own. How is Canon able to do this? It should have the same artifacts as the Sony, right? But it doesn't. It looks fantastic. I am in awe.

[Tony Tibbetts]

Can someone point out which clips are specifically in 24f mode, I'm not really that interested in the others and I don't really wanna waste time and bandwidth on downloading clips that aren't 24f

[Kaku Ito]

Originally Posted by Tony Tibbetts

Can someone point out which clips are specifically in 24f mode, I'm not really that interested in the others and I don't really wanna waste time and bandwidth on downloading clips that aren't 24f

The ones says 60 or 30 is not 24 frame mode, others are all 24 frame mode (since members here asked for 24 frame mode).
What kind of camera work do you wanna evaluate? Maybe I or other members can recommend a right file for you.

[Stephen L. Noe]

Low light my friend, low light. Try to shoot some stuff just under street light (around 60 lux). Low light seems to be the acid test for 1/3 high def cameras.

It seems the H1 will suffer the same smearing the JVC does. I've noticed it in the CA shot. I'm looking for a similar split screen in the H1 as well at low light. Let's see how it performs 24F @ 40-60 lux.

Kaku Ito, are you using the ND filter on the outside shots normally? The reason I bring this ups is because the waveform shows the highligts pegged for every outside shot.

Click here to see what I mean

I think all the 1/3" HD/HDV cameras are going to suffer in some way, merely because of the size of the chips. Pick your poison.

[Kaku Ito]

Originally Posted by Stephen L. Noe

Low light my friend, low light. Try to shoot some stuff just under street light (around 60 lux). Low light seems to be the acid test for 1/3 high def cameras.

It seems the H1 will suffer the same smearing the JVC does. I've noticed it in the CA shot. I'm looking for a similar split screen in the H1 as well at low light. Let's see how it performs 24F @ 40-60 lux.

I think all the 1/3" HD/HDV cameras are going to suffer in some way, merely because of the size of the chips. Pick your poison.

Immediate feeling I had with this camera was that it is a lot more sensitive than FX1 and HC1. You can check out all of the night clips (the ones named with "nite" word) for low light condition.

About ND filter, I use it according to the situation. In this case I wanted to get more of the things in the shade to be seen, so I might have made it little brighter. or anything to do with 0IR setting?

[Stephen L. Noe]

Man these forums are quick. Check out my edited entry just above yours amigo.

best,