DV Info Net

DV Info Net (https://www.dvinfo.net/forum/)
-   Canon XL H Series HDV Camcorders (https://www.dvinfo.net/forum/canon-xl-h-series-hdv-camcorders/)
-   -   Kaku Ito's XL H1 video clips now available for download (https://www.dvinfo.net/forum/canon-xl-h-series-hdv-camcorders/52060-kaku-itos-xl-h1-video-clips-now-available-download.html)

Nate Weaver October 2nd, 2005 01:48 AM

Quote:

Originally Posted by Lauri Kettunen
In the upper right corner the wall of the building which is in light has a magenta tone. So, I wonder, whether this is a white balance issue?

That's classic chromatic aberration. One edge will have magenta ghost, the other will be green.

Quote:

Originally Posted by Lauri Kettunen
Obviously the heavy compression must have some side effects. No free lunches available.

Sure. Do some super shaky handheld work and take a close look at how the picture will soften up during fast movement. That's the easiest viewable artifact of HDV under stress.

Barry Green October 2nd, 2005 01:56 AM

Quote:

Originally Posted by Thomas Smet
Barry do you know or think tapes in 60i would work fine in a current other 1080i deck/camera? I can see why the 24f and 30f may not work because of the flags but what about straight everyday 60i?

Well, I'm tempted to say that I can't imagine how the XL H1's 60i tapes wouldn't play in the Sony gear... but then again, Canon appears to have deviated from the spec by including four channels of audio (each channel using 96kbps, for a total added-together rate of 384kbps). So will that play back in the Sony? I don't know. I would be very surprised if 60i tapes from the Canon don't play in the Sony.

Quote:

Kaku maybe if you get a chance you can see if tapes recorded in 60i playback from your HC1?
Excellent idea, and yes, Kaku if you could test that we would be most appreciative! Please test 30F also, if you would.

Thomas Smet October 2nd, 2005 02:02 AM

One thing I will say whatever CANON is doing to get to 24p it does it very well. Without looking at the lines of resolution or detail the images are very clear with no sign of deinterlacing or aliased edges. I tried checking many different angles of lines with different levels of motion and so far I haven't been able to find any aliased edges due to an interpolated field of data.

While some of the early footage from the H1 may look like video the very first footage from the HD100 in my opinion at least looked much much worse. It wasn't until the HD100 was really pushed that we saw it's potential.

My only concern at this point is to see what is happening to the chroma during CANON's conversion to 24f/30f. If it is like the frame movie mode or even just 540 lines duplicated we kind of end up with 4:2:0 all over again. Of course this doesn't matter on HDV tape but it does matter for the HD-SDI output where we would have been able to get 4:2:2.

In this case it might actually be better to use the component uncompressed output from the JVC HD-100. Even though it is 720p at least you know no tricks were done to get it to 720p and you will get a true 4:2:2

Chris Hurd October 2nd, 2005 05:24 AM

Quote:

Originally Posted by Stephen van Vuuren
Interesting - I think Barry Green is absolutely right about his 24F theory. Objects in motion seem unaturally soft compared to still background.

I'll pretty much always defer to Barry Green, but isn't this a hallmark of HDV, not the frame rate? Canon has something similar to JVC's "motion smoothing" in the XL H1. They're not calling it motion smoothing, but something like it is in there, no matter what the frame rate. It's just more apparent at slower frame rates.

Chris Hurd October 2nd, 2005 05:33 AM

By the way folks with regard to any more clips from Kaku today, I will have those posted and linked for you as soon as I return this afternoon. Going out on a shoot right now and will be back in about eight or ten hours or so.

Kaku Ito October 2nd, 2005 05:33 AM

quick note
 
I labeled the clips by the display shows what mode the camera was playing. I was relying that so if that was not displaying correctly, they could be labeled wrong. I would double check.

Kaku Ito October 2nd, 2005 07:02 AM

Since I got an another obligation, I had to shoot mountainbike race in 60i format, but thanks to my new employee who is also mountainbike lover helped me to shoot FX1 with the same angle. So these clips are coming after cutting and uploading 24f clips for filmmaking target member here. I managed to shoot a scenery of a sunny bay around 4 o'clock with no gamma curve, cine 1 curve and cine 2 curve. They are already being uploaded so you will see them as soon as Chris comes back from shooting. If these clips appear to be 30 or 29.97, I don't know what is happening. Maybe something is happening when Lumiere HD is capturing? I have to ask Frederic about that.

I also shot various scenes with Cinegamma 2 on. I will make short clips for that too. I'm not sure I'm going to be able to shoot the chart and green back since I have to capture all of the footage before I return the camera.
I will do my best to try those. Maybe have my employees work on them tomorrow morning before we return the cam. But people already see the aberration happening with my bike clips so it seems to be clear that the lense is not quite comparable for real HD because of the price? But what I see is that comparing the bikeseq files with XL H1 and FX1, I think XL H1 looks the best when I zoom all the way in at x20. And low light ability is clearly better on XL H1. What I really wanna see is the compression characteristics between the two.FX1 and HC1 are not capable of shooting mountainbiking race (cross race and downhill) without compression artifacts, I need one camera that doesn't in the same format. If XL H1 is excellent, then I would sell my FX1 and try to buy XL H1 and use HC1 for more handy use.

baygamma files are uploaded now (Chris has to put them listed for you), so I will search good portion of my footage for different situation. Be back soon.

Kaku Ito October 2nd, 2005 08:56 AM

I was discussing with Frederic Haubrich and all of the XL H1's 24 frame clips will be 24 frame simulation and it is based on 1080/60i format. So if you open them in whatever the softwareto playback, it might say 29.97. Is that the case?

Soroush Shahrokni October 2nd, 2005 09:01 AM

Quote:

Originally Posted by Nate Weaver
Alright. View, in VLC back to back, if you will:

http://homepage.mac.com/nweaver/.Movies/XLH1-24p-1.m2t
http://homepage.mac.com/nweaver/.Mov...ywoodBlvd1.m2t

The XLH clip is in 24.

Thx for posting the 24F clip Nate. I had seen your Hollywood clips before and now that I saw it back to back with the 24F clip I must say that Im more convinced that 24F doesnt give the same feeling as real 24P. I really dont know whats missing bc it looks much better than CF, but it still doesnt look like 24P to me!

Kaku, thanks for uploading new footage. Cant wait to see them!

Stephen van Vuuren October 2nd, 2005 09:14 AM

Quote:

Originally Posted by Nate Weaver
It appears to me that Canon did the same trick, using flags. It has nothing to do with how the Canon got those 24 frames from the CCDs....that's another issue entirely.

Where are you seeing repeat flags? I'm not seeing that. The 24F has twenty four frames, the 30F frames etc.

Kaku Ito October 2nd, 2005 09:25 AM

24frame portion of XLH1 tape did not play on HC1. It said incompatible format. However, 60i portion of the tape played fine.

Stephen van Vuuren October 2nd, 2005 09:26 AM

Quote:

Originally Posted by Chris Hurd
I'll pretty much always defer to Barry Green, but isn't this a hallmark of HDV, not the frame rate? Canon has something similar to JVC's "motion smoothing" in the XL H1. They're not calling it motion smoothing, but something like it is in there, no matter what the frame rate. It's just more apparent at slower frame rates.

I thought JVC had motion smoothing to cover for the fact that it did not shoot interlaced. Why would Canon need this since it shoots interlaced just fine?

Kaku Ito October 2nd, 2005 09:34 AM

Quote:

Originally Posted by Soroush Shahrokni
Thx for posting the 24F clip Nate. I had seen your Hollywood clips before and now that I saw it back to back with the 24F clip I must say that Im more convinced that 24F doesnt give the same feeling as real 24P. I really dont know whats missing bc it looks much better than CF, but it still doesnt look like 24P to me!

Kaku, thanks for uploading new footage. Cant wait to see them!

19 more 24 frame(emulation) files are uploaded to the server now including three gamma settings on ocean scenery. It seems to be no sense to go without the cinegamma setting, so I shot all of the 24 frame mode clips with "cine2" gamma setting and black parameter set to "stretched". I also did 60i comparison between XLH1 and FX1 but since I can edit that later (without the cam), I will work on that later.

Just wait for Chris to come home and feed you all.

Soroush Shahrokni October 2nd, 2005 09:59 AM

Kaku, arigato gozaimashita!

Lauri Kettunen October 2nd, 2005 10:54 AM

Quote:

Originally Posted by Nate Weaver
That's classic chromatic aberration. One edge will have magenta ghost, the other will be green.

Ok Nate, I'm now with you: Did not spot the green ghost at the first glance. Obviously, a simultaneous green and magenta ghost cannot be removed by adjusting the white balance.

I've been also thinking of this deck issue, for, so far, more involved editing has called for a deck. But if new portable hard disk devices for storage appear, then, perhaps, it's better to store both on tape and hard disk, and use the tape only as a backup. (Just read yesterday from the newspaper that these new dvd discs will be able to store up to 100 GB of data.)

My more immediate concern is, will my editing software (Premiere Pro) be able to import all the footages from XL H1. At least, it seems that currently the Premiere Pro HDV mode does not recognize 25P (25F).


All times are GMT -6. The time now is 04:42 AM.

DV Info Net -- Real Names, Real People, Real Info!
1998-2024 The Digital Video Information Network