I hope my plan works on July 7th. Not for sure tho. But I will try.
You won't be seeing the same green bike anymore. Just got stolen last night.

Now I can provide you folks with better choices of video formats since FCP5 is here.

Wow - you had a bike stolen in Tokyo?


its probably outside Shinjuku station :)

Hi John Jay,

No, it happened in Shibuya.

Great! I'm waiting impatiently for your review/comparison. :)

Regards

I hope my plan works on July 7th. Not for sure tho. But I will try.
You won't be seeing the same green bike anymore. Just got stolen last night.

Now I can provide you folks with better choices of video formats since FCP5 is here.


A danged shame with the bike theft. Hopefully you can film an event for someone and earn enough money to replace it with a nice upgrade!

From page 87 of the HC1 manual, it says "subjects passing by the frame might appear crooked". This suggests that the CMOS sensor is using a rolling shutter rathar than a global shutter. If true, it would mean that Sony is getting more light sensitivity out of the available die space on the sensor by not having an extra shielded sensor area dedicated for a virtual shutter. The drawback is that objects in rapid horizontal motion show up tilted. The faster their CMOS readout speed can be, the less tilted it will be overall, but still it would be present. This same effect, much more pronounced, can be seen with cheap webcams when there is horizontal motion in the scene. In that scenario you can clearly see the scanning effect being used by the sensor.

Anyway, if you have time Kaku, I'd love to see an .M2T of an outdoor scene with the camera set to a high shutter speed, and something blocky (like a large truck) driving by rapidly. Or rapid panning on a scene with strong vertical lines. Examining field by field (cutting each frame into their individual fields) should show what appears to be a slanted edge. If the object is moving rapidly to the right, it should look like it slants slightly to the left!

Even if you don't have time for this specific test, I'm still very eager to check out the M2Ts you'll be creating. And thank you very, VERY much for your willingness to share them.

-Lorin
HDV Forever! (http://hdvforever.com/hdv/)

Kaku, we've hosted your FX1 clips before and we'll be happy to do it again for your HC1 clips. Just let me know when you're ready and I'll give you the access info for uploading. Looking forward to it!

Kaku, again I'll say that you've done AMAZING work on our behalf in North America - last summer you were invaluable helping me choose GS400 and then HDR-FX1... can't wait for your info!

Mark in Toronto

Kaku

I've noticed that the Aperture on the HC1 goes from F1.8-2.1 which is hardly anything compared to the F1.6-F11 of the FX1. This would indicate that the performance of the sensor might be so good as to not require much stepping down in Bright light. But also an advantage, even though you won't get large DOF, the effect of AE (and manual adjustment) on DOF will be reduced, and mostly effect gain (once shutter is locked). Along with locking down shutter speed, this gives a much better control of look of the video. Something well worth testing against the FX1.

Re-edit: I should add, not enough control to shoot cinema quality, of course, only doco etc.

Here are some other things that might be worth testing out:

http://www.dvinfo.net/conf/showthread.php?t=47018
http://www.dvinfo.net/conf/showthread.php?t=47019
http://www.dvinfo.net/conf/showthread.php?t=47020

You people are so great and on top of things on what I should be doing and since what we (me and my company) can learn from it is so great, we decided to purchase to cam for testing (we can probably sell it right away).

With the help of Pro Video Station Shinjuku, I will be able to receive the cam on 6th, which is this Wednesday.

Great -- glad to hear it, Kaku! Let me know when you're ready to upload some clips,

I've noticed that the Aperture on the HC1 goes from F1.8-2.1 which is hardly anything compared to the F1.6-F11 of the FX1.Er... Wayne... hold on there just a moment.

When you see f/1.8 to f/2.1 on the HC1, that's not the full aperture range of the lens by any means. It's simply the maximum apertures at each end of the zoom range. The maximum aperture at full wide is f/1.8. The maximum aperture at full telephoto is f/2.1. I couldn't readily find what the minimum aperture setting is before the iris closes, but most likely it's going to be f/11. Hope this helps,

Okay, Chris and others that gave me testing suggestions.

Let's build the most valuable consumer report (opinion) here!!
I'm very concern about how Sony is skipping on cultural value with music and video lately, so I want them to be aware that the consumer expectation on media quality is higher than what they think. At the same time, they develop and manufacture good technology, too, so I want them to be better in taking good care of media culture. Our study and conclusions will greatly influence Sony.

...consumer expectation on media quality is higher than what they think.

I think you're right. I hope Sony takes note of how interested we are in the details. Quality is very important to us. We're definitely happy to learn from all the tests you're preparing for.

Based on a few .M2Ts that were released to the Internet this morning from another site, it looks like the SteadyShot feature will occasionally stretch frames vertically to try to compensate for motion. Not a very good solution to stabilize the image! Ends up with a slightly "pulsating" image, and also limits a little the effectiveness of the MPEG2 compression. But that's really my only gripe with the footage. I'm hoping that this type of single-frame stretching of the image can be avoided by turning the SteadyShot off.

By the way, I'll be flying into Okinawa prefecture on Sunday, with a four-hour layover in Narita. If you know of a good place near either Narita or in Naha to buy an HC1 for around 160,000 yen, I'd be interested. Do you know if any of these three shops are reputable? They sure have great pricing.

http://www.tantan.co.jp/detail/HDR-HC1 - 151,400 ¥

http://www.tokado.jp/gekiyasu/video_body.htm#MARK1 - 158,300 ¥

http://www.bidders.co.jp/pitem/50723344 - 155,757 ¥

Thanks!

-Lorin

Lorin,

Four-hour layover is not much to do since Narita is far from Tokyo.

You should make arrangement now with people in Okinawa that you can see in person to preorder for you with one of these company, or the shop I'm buying from. I can make sure that they help you.

Are camera's cheap to buy in Japan ?I will be over there in September and was wondering what it costs for an XL2 or the new JVC GY-HD101E .Also are PAL versions available? Joe

Joe, well, maybe it is right now, since yen is very weak at this time.
The dealer I deal with is selling XL2 (body only, without the lens) for 321,300 yen (only supply lasts). You probably find more information in the XL2 thread better than here.

Pro Video Station Shinjuku (http://www.pv-station.com/product_info.php?cPath=2020&products_id=1454)

Chris, OK, I thought that sounded a bit too good to be true :|( no blush cheek smiley available for some reason ;) . I checked over the web and found the same figure, and even f1.6:f11 in comparison for the FX1. So I thought maybe they were trying to strangely handicap the camera by removing larger DOF's. Silly me ;) Still, approx half a stop loss for telephoto is good.

Okay, my staff emailed me (I was out to give lecture at a music college) and the cam is in.

I will make few clips available on midnight.

Chris, you can copy the files to make them available at your ftp. I will close mine after you copy them.

Now, I'm off to the same Gap and the street to shoot the same angles and movement in Harajuku.

Thanks Kaku -- I'll let you (and everybody else) know when we have them.

Now I'm having the same problem that Frederic had, can't capture with FCP5 and 10.4.1. Do I have to reinstall the whole OS?
FCExpress is capturing okay, but it will be AIC that way.

I guess I'm going to install 10.4 into my G-Tech G-RAID.

Because of this, it will take few more hours to start providing the files.

After reviewing other M2Ts out on the 'net, some have a vertical stretching type artefact. I think this stems from the HC1's CMOS sensor having a rolling shutter. If so then it must be scanning the sensor pretty slowly! Now I'm curious to see footage that has some vertical motion, across different shutter speeds. I think we may have a significant rolling shutter issue on our hands here.

Technically the faster you can read a sensor with no shutter, the better it should be. But from what I'm seeing here it may only be scanning at the rate of the shutter speed.

In addition to vertical motion uncovering this issue, another way it would show up is when a flash goes off in a frame. Normally on film or a CCD system with a shutter, if you have a quick flash such as with flash photography, lightning, an explosion, or a strobe, then the whole field is affected simultaneously. Those that have done frame by frame editing of events are well aware of what it looks like. One frame is normal, the next is brighter from the flash, and then after that it's back to normal again. But with a rolling shutter it will be a horizontal band, and not always affect the entire field. It would at least occasionally (and perhaps frequently) only show up on a part of the total field! Perhaps just a few lines, perhaps a broad section. So most of the image would be normal, but a portion would have a brighter band in the middle.

Altogether this is probably a new beast we're dealing with here. Good deshaking algorithms could compensate for some of the motion effects. But subjects travelling horizontally, especially moving quickly, would be tough to correct for.

(Anyone else out there technically inclined and understand what in the world I'm talking about?)

-Lorin

Lorin,

The footage I just edited, was shot with shutter speed at 60 and it include the video of water and coffee going down from espresso machine.

Now I'm waiting for chris to provide me the login procedure to upload the edited m2t file.

I will do more test tomorrow on something going up and down (maybe cars going by with camera set sideways) at different shutter speeds.

Here's the first clip for now.

This footage was shot with shutter speed "60", automatic f-stop and gain. It was shot early evening. Atuomatic white balance.

Probably it takes awhile for file transfer to complete.

dotMac filesharing page (http://homepage.mac.com/kakuito/filesharing/FileSharing215.html).

P.S. I could not do the comparison today (I was alone shooting video), but I will ask my wife to help me tomorrow.

Thank you Kaku for the clip. <wow> It's almost 300 MB - the longest clip from HC1 available. Unfortunately I will have to wait till tomorrow to watch it - it's at night now and it downloads "only" at 20kB/s.

Thank you

I've downloaded the clip, but i'm only getting the audio. I'm using both the VLC player and WM, neither work. I could view Kerr Cook's clips just fine, but not these.

Any suggestions?

I have the same problem

I left it as QuickTime HDV format and maybe that is why. How about downloading QuickTime 7? Will that work?
Anyway, I'm converting the file to "HDV1060i format" in "Compressor 2", so that should be more native?

I left it as QuickTime HDV format and maybe that is why. How about downloading QuickTime 7?

Not for me. I've tried it on two different PC's with Quicktime 7.0 preview with no luck. I've been searching for codecs all night, but just cant seem to find one that works with this particular HDV file.

-Mike

Mike,
Thank you for trying. It takes hours to transfer to dotMac acount and that goes same to you to download. I asked my company web administrator to setup download page in my site, so I can be more flexible in creating formats for Windows users.

For Windows users, should not download this file. I will capture the clip using LumiereHD (just like the first time) and post it as native m2t.

I wonder how mac users are doing.

i cant play it too

Hello all.

I can play the clip fine on Windows XP with Media Player Classic. I am using version 6.4.8.2 so that and newer should work.
It uses the built-in mpeg2 codec it seems.
Here is a link to it on Sourceforge: http://sourceforge.net/project/showfiles.php?group_id=82303&package_id=84358

Best, Magnus.

Here's the first clip for now.

This footage was shot with shutter speed "60", automatic f-stop and gain. It was shot early evening. Atuomatic white balance.
Just like some of the other guys, I get video but no audio. And very nice at that! The focus seeks a bit when it's zoomed up on the grounds. Amazing demo of the latitude, even with a fairly bright background and dim foreground the background doesn't get washed out. EIP works very well in that regard. 14 bits is enough to be impressive. Gets kinda grainy at +15db... But still usable. Definitely not a low-light cam.

I think the red truck that passes by may be a good enough subject to examine the rolling shutter artifact with. I'll start picking it apart frame by frame.

For those that can't view it yet, note that even though the file extension is M2T, the content is actually recompressed into a program stream at 26.3 Mbps. So not too much loss altogether. Still a standard 15GOP. I was able to play it back full speed on my laptop, which has a 2 GHz Pentium M.

Thanks very much for the clip! Time for me to put it under a magnifying glass.

-Lorin

...the content is actually recompressed into a program stream at 26.3 Mbps.
And also put into a QuickTime wrapper with strange indexing! No wonder some software didn't understand it that well the first time 'round.

-Lorin

The video played fine on my XP with Real Media Player. This machine I use for the internet is only a Athlon XP 1800 and it played fine. It was a little jerky but it still played.

The video looks very nice so far. Not bad for a single chip camera. They sure have come a long way. I will be some image tests later tonight. Thank you for the video. You did a great thing for all of us.

One thing I did do was change the file extension to ".mpg" I can now even use the file in After Effects.

I think this is QuickTime 7 thing. But I knew you folk would figure out some other ways to playback.

I mastered this clip to tape and try to capture it with LumiereHD. Than will capture the footage as simple m2t, but can anybody figure out what is the best way to convert the edited clip from FCP5 to m2t that both mac and window users can playback? I'm busy shooting and editing so, if anyone can help, that would really save time and effort. While someone will be figuring that out, I would be editing more clips.

Also, we went out to shoot some cars going by on a street with both HC1 and FX1 at the same time with all of the shutter speed from 10,000 through 30. Lorin, which shutter speed footage would be needed to compare (I'm going to go ahead and do 10,000 and 60 for now which are probably mandatory anyway)?

AND, the focus is rather slow in dark places (with low contrast I guess). With its small LCD and the view finder, I was having harder time focusing manually comparing to FX1 for sure. The focus/zoom ring feels fine, not as good as AG-DVC30 felt, but better than NV-GS400 for sure. For how fucus/zoom ring works, AG-DVC30 still feels the best to me.

i change it to .mpg and played with windvd 7 and it worked but only video - no sound. With media player classic the opposite only sound - no video.

Please Kuku Ito
can you take video with hc-1 while walking? I always record like that with a small-cheap steadycam.
thank you very much

Yiannis
Greece

The footage looks real nice, but as suggested before, some raw M2T files would be nice to check out. Since the file you provided is edited and recompressed twice it is hard to judge the quality straight from the camera. From what I see, it looks pretty damn good.

Thank you for your effort.

Kyle,
It is more trouble for me to boot up something other than Final Cut Pro 5 right now, but I will come back and provide the raw files later. Also, in the beginning I need to show credit for what I do and what people helped me for and raw files can't include these things.

Yiannis,
I actually noticed the stabilizer worked pretty good for that purpose. The one I shot last night is too dark, so I will do something else for you.

Lorin,
I posted the footage file with car passing by (only side way for now) which are shot with both FX1 and HC1. Can you look at that? It is only 20 sec. footage, so the file is smaller and please tell me what we can do further.

xtream.ne.jp download page link (http://www.xtream.ne.jp/content/view/168/33/)

Please register at my site and leave comment for me there, too. I look like an idot just talking to myself there. I really appreciated that Chris, Mark and few other people left messages at my blog when I provided the FX1 files last year. I'm hoping to do active discussion here and sort out my idea there at my site, so both my site and hdvinfo gain from it.

The automatic adjustments seem to lead to little on the brighter side during the night shooting. I kept going to the brightness adjustment to lower the brightness to make it look close to what I see with my own eyes (hoping that the gain would go lower and gets less grainy).

which shutter speed footage would be needed to compare (I'm going to go ahead and do 10,000 and 60 for now which are probably mandatory anyway)?

Those two speeds should work fine. For the cars going by test the camera should be used normally, so cars go by horizontally. Not with the camera tilted on its side.

The rolling shutter effect should also be visible when filming with some varying vertical motion, like what Yiannis had asked for, some hand-held walking shots. Just walking normally would be fine. And for a more extreme test, use a high shutter speed and move the camera with a short motion up and down more quickly on a detailed scene.

Thanks so much, Kaku!

-Lorin

Lorin,

So the comparison file should be good for your testing. They were shot without the stabilizer, but that is not concern anymore on the vertical tilt, right?

It is already dark here in Japan, so I can't shoot good footage with vertical movement, but I will do that tomorrow.

Also, someone was asking for a human skin footage on another thread. I shot video of a little girl playing around today, so I will post it sometime tonight.

They were shot without the stabilizer, but that is not concern anymore on the vertical tilt, right?
Super SteadyShot hasn't been ruled out as the main culprit, but after pondering on the effect some more it has taken a back seat. So what you've captured is definitely still a valid test.

Thank you again so much!

-Lorin

Thanks very much for the clip! Time for me to put it under a magnifying glass.
Okay, I've been checking it out, and here's a portion of one of the frames giving evidence that the sensor uses a rolling shutter:

http://hdvforever.com/hdv/truckslant.jpg

Notice the red truck looks like it's slanted backwards. I would have to expect that's not what trucks in Japan really look like! (The back of the truck would be fairly straight, and fairly vertical.) Although subtle, the rolling shutter effect is present.

This is then also the most likely culprit for the vertical stretching phenomenon I saw in some other clips. Hard to see when it's being played back full-speed, but evident when going through frame by frame.

When clips taken with a higher shutter speed are made available, the effect will become more obvious. But perhaps it will still be an acceptable compromise for most people.

Still can't believe the broad latitude this camera offers. CMOS really trumps CCD in that area. Very little need for a neutral density filter.

-Lorin

About rolling shutter effect, some slight vertical image breezing on some frames with steady shot. Problems, if they exist are so slight, no one has really complained about.

Slight rolling shutter effect exists in film cameras too.

We have preproduction model, tested on another site, was supposed have some extremely minor image breathing.

Even if was rolling shutter and breathing existed in production model, would it effect anybody's purchase decision? Extremely unlikey. Camera's superior overall performance is what natters. No camera is perfect.

As to rolling shutter with Sony CMOS, any CMOS, Sony have 3 CMOS 3MP DV camera. Haven't heard anyone raising issue about rollig shutter issue. Even if rolling shutter existed, to me is non issue, rather nothing to really worry about. The camera has no competition. There is no other camera that size, price that come even close in performance. Still is nice to know.

Steady Shot can bring lot more image degradation than what described. That's why is switchable and is hardly ever used in true pro productions.

Radek

So far, the cam is more than satisfactory for the price and ease of carrying around. I won't even have problem to set it on my helmet to do onboard video (not sure if the MPEG scheme can hold on with all of the vibration) as far as the size and weight you are talking about.

Tilting problem seems to be minor. I have more clips to share tomorrow with little girl playing in a small river in the park, a dog, and almost famous Harajuku street descending (and GAP). Now Frederic Haubrich helped me with migrating my LumiereHD to the new machine, I will be able to provide the raw clips, too. Then we can compare how well or bad FCP5 keeps the original fidelity on HDV.

By the way, I added m2t format file (converted from FCP5 HDV format) on my site, so people had problems playing back can download that version to enjoy. <Click here> (http://www.xtream.ne.jp/component/option,com_docman/task,cat_view/gid,54/Itemid,68/)

By the way, I added m2t format file (converted from FCP5 HDV format) on my site, so people had problems playing back can download that version to enjoy. <Click here> (http://www.xtream.ne.jp/component/option,com_docman/task,cat_view/gid,54/Itemid,68/)

Kaku, that link isn't working properly. It states "You must login to access the document section."

Edit: Nevermind. I decided to just go through the registration process and created an account for myself.

I'm not sure, but it looks like HC1 has incredible dynamic range. In Kaku's clip the gain is above 9db, grain is clearly visible inside car (e.g. at the face), but also the portion of picture where is view at outside (tree, other car) is almost perfect - without grain and not overexpose. Maybe I'm missing something, but I don't think any consumer/prosumer CCD camcorder would be able to capture such a high dynamic range. Is Sony leaflet about CMOS advantages true, what do you think of it?