|
|
|
|
|
April 13th, 2007, 09:08 AM
|
#181 |
|
New Boot
Join Date: Mar 2007
Location: Jacksonville, Florida, USA
Posts: 24
|
More sample footage
I am posting links to low light footage I shot yesterday. The files will be available for 7 days or 100 downloads.
The light was provided by two 150W halogen lamps pointed at high white/slightly gray ceiling, at my desk level there was just barely enough light to read a book. But it was evenly distributed. Camcorder on tripod, OIS off, manual focus, auto WB, F1.8, 1/60 sec exposure (except where noted otherwise). Big red ball bouncing, note how the gain changes in the background when darker ball fills up the frame: http://www.yousendit.com/download/Ql...QzNiV3cwTVE9PQ Yellow ball bouncing 1/60 sec: http://www.yousendit.com/download/Ql...SWVvQnMwTVE9PQ Yellow ball bouncing 1/250 sec (gain was at max by itself): http://www.yousendit.com/download/Ql...SU9nYU0wTVE9PQ White balance on auto (a white card entering the scene): http://www.yousendit.com/download/Ql...Tk04NVUwTVE9PQ WB manual first balanced on white card, then shot: http://download.yousendit.com/BAC53FA05A8F318E Later I will post some samples showing the CA. |
|
|
|
April 13th, 2007, 09:15 AM
|
#182 |
|
New Boot
Join Date: Apr 2007
Location: Madison, WI
Posts: 17
|
I'm obviously over my head here, so I'm just going to keep my mouth shut and listen to the experts and hope to learn something :)
Erik |
|
|
|
|
April 13th, 2007, 09:35 AM
|
#183 |
|
New Boot
Join Date: Mar 2007
Location: Jacksonville, Florida, USA
Posts: 24
|
Yousendit limit extended
I am sorry, but I did not realize that beyond limitations of only 7 days and 100 downloads there is also a 1 GB download limit on yousendit.com for free accounts... What is the point of having a free service that does not work?
Sorry guys, I will try to find a more reasonable service and repost these clips. I have more clips to share with you, so you could judge by yourself whether this is a camcorder for you... Also want to add another observation. The file numbering is very odd. It goes from lets say MOV030, MOV031, ... MOV039, MOV03A, MOV03B, ..., MOV03F, MOV040, MOV041, etc... So when you copy the files to PC and sort by name - your clips are not in chronological order. Why JVC did it this way, I have no idea, but it inconvenient. There is a small 1KB file with extension *.MOI for each *.TOD file, presumably storing shot parameters, but its content cannot be viewed. The bundled SW does not show the shot parameters either or maybe I just did not find how to display them yet. The cam stores the list of clips on cam in a XML file, which can be opened in a browser, but it does not have shot parameters either. |
|
|
|
|
April 13th, 2007, 10:07 AM
|
#184 | |
|
Regular Crew
Join Date: Nov 2006
Location: MO
Posts: 41
|
Quote:
If they include a player and the image is I instead of P then the case is closed. I have downloaded the first 3 but put only one on my server here: http://s87224782.onlinehome.us/MOV022.TOD |
|
|
|
|
|
April 13th, 2007, 12:09 PM
|
#185 |
|
New Boot
Join Date: Mar 2007
Location: Jacksonville, Florida, USA
Posts: 24
|
Found a new file sharing site
Because my yousendit free account download limit was exceeded in only few hours I found another site and
this does not want you to register and is much faster too. Unfortunately it only gives 20 downloads per file valid for 5 days. So if it runs out, it runs out. I cannot keep uploading the same files all day long. Sorry guys. Big red ball bouncing, note how the gain changes in the background when darker ball fills up the frame: http://www.transferbigfiles.com/Down...9-de591b1c3901 Yellow ball bouncing 1/60 sec: http://www.transferbigfiles.com/Down...f-9e02b37aef59 Yellow ball bouncing 1/250 sec (gain was at max by itself): http://www.transferbigfiles.com/Down...b-f989f5a61117 White balance on auto (a white card entering the scene): http://www.transferbigfiles.com/Down...3-adb5c315fe0a WB manual first balanced on white card, then shot: ...coming... Later I will repost the videos I posted yesterday and also new ones. Good luck. If someone has an FTP they can share I am willing to upload everything one more time. |
|
|
|
|
April 13th, 2007, 09:58 PM
|
#186 |
|
New Boot
Join Date: Mar 2007
Location: Jacksonville, Florida, USA
Posts: 24
|
More samples
WB manual first balanced on white card, then shot:
http://www.transferbigfiles.com/Down...3-adb5c315fe0a Kid on bike, 1/250 sec, manual focus, tripod. Note light streaks from sun reflections on helmet. This is the video that I was trying to render as a progressive video (not deinterlacing it, but "properly" combining the two half frames from interlaced file). If anybody finds a solution for playback and decoding of this file as progressive - PLEASE POST YOUR SOLUTION BACK HERE! I tried the bundled software and Vegas 7 so far no luck... http://www.transferbigfiles.com/Down...2-6d9e70b8bd99 Wind blowing through tree, lots of moving fine details: http://www.transferbigfiles.com/Down...2-cf892d2f14b5 CA in worth case scenario - thin branches against the sky: http://www.transferbigfiles.com/Down...a-00c8999ba57d To confirm the battery life issue. Not only it is less than 2 hours (I am not sure how much, but closer to one hour), but it also takes very long time to recharge. From fully discharged (at least the camcorder stopped functioning) to fully charged (when the LED stopped blinking) it needed 2 hours 45 minutes. The battery remained cold. So an external charger and a second (and third) battery would be a must. When you shoot from power supply it does not charge the battery either... Another unpleasant surprise, the sensor exhibits charge leaking from grossly overexposed pixels (streaks). When the electron well of pixel becomes full it leaks along the column of pixels. This is a sign of cheap CCDs. I am more and more disappointed... The tripod mounting hole is way back and to the left of the center of gravity of cam. Not a major thing, but something to know about. |
|
|
|
|
April 14th, 2007, 12:02 PM
|
#187 |
|
New Boot
Join Date: Mar 2007
Location: Jacksonville, Florida, USA
Posts: 24
|
An easy way to find out whether you cam captures progressive or interlaced video
I have good news for all of you guys! I have a proof that JVC HD7 camcorder is in fact progressive!
I was scratching my head how to prove it with 100% certainty without relying on any software and figured it out this morning. This method is extremely simple and 100% reliable, does not require any special processing software or any knowledge of computers. It can be used with any camcorder regardless of recording format (DV, HDV, HD) or region settings (NTSC, PAL). You are welcome to adopt it, but please give me the credit for inventing this method. I may even consider patenting it, why not? People patent totally obvious common things and this is not so trivial... What you need for test: ======================= You need tripod to stabilize the camcorder you want to test, a flash and a moving object (a ball or kids car or something you can move in front of the camcorder). The background should be of substantially different reflectiveness (brightness) than the object. Lock you cam on tripod, focus to objects distance and set focus on manual. Adjust your exposure speed to 1/30 sec in case of 30 frames per sec video and to 1/25 sec in case of 25 or 24 frames/sec video. This will ensure that there are not gaps in between the frames when the light is not accumulated by sensor. With short exposure times you get less motion blur, but resist the temptation, the motion blur is not at all important in this test. The reason we want to capture light without any gaps is because the flash light burst lasts only for for 1/200 of second and if it flashes in between two frames you will not see it. Adjust your gain or aperture (whichever you cam lets you adjust) such, that the overall brightness of the video is low, in other words the frame is dark. The test itself: ================ Start capturing video without the object in frame, move the object through the frame at the distance you focused at, fire your flash pointing at the object when it is in the middle of the frame, stop the recording. Evaluation of results: ====================== Now I have to provide a little explanation for evaluation of results of test and what you would expect to see. I apologize for the next two or three paragraphs, I know many of you understand the "mechanics" of video recordings, but some don't and this is for their benefit. In interlaced movie, odd lines captured in the first half of 1/30 sec exposure, even lines captured in the second half of 1/30 sec exposure. The integration time (the time the sensor accumulates photons and generates electrons) is 1/60 sec. All moving object or stationary object while cam pans will show the "comb" on their edges, as there is 1/60 sec time difference between the odd and even lines captured. ################################### .......################################### ################################### .......################################### ################################### .......################################### ################################### .......################################### ################################### .......################################### ################################### .......################################### In progressive movie, full frame, both odd and even lines captured during the 1/30 sec exposure. The integration time for each row is 1/60 sec. The moving objects or stationary when camcorder pans, do not show comb effect. ################################### ################################### ################################### ################################### ################################### ################################### ################################### ################################### ################################### ################################### ################################### ################################### Depending on the recording format, both progressive and interlaced movie can be stored in progressive or interlaced way. It is up to the player or editing software to properly recognize the recording format and reconstruct the original video information. And this is the most difficult part - how can you be sure you video editing suit renders you movie properly? You can not! But with my test you can now! So here is how you evaluate the test results and what you can expect to see: ============================================================================ Interlaced movie: ================= 1. dark frame 2. dark frame 3. flash 4. dark frame Here is the frame with the flash: ################################### ...........dark line................. ################################### ...........dark line................. ################################### ...........dark line................. ################################### ...........dark line................. ################################### ...........dark line................. ################################### ...........dark line................. ################################### Every other line will be dark (does not matter if those are the odd or even lines) and the frame just before and just after it will be totally dark. This proves that the 1/200 burst of light was captured in one half of one frame - an indication of interlaced capture. However it is recorded it remains interlaced! Progressive frame recorded progressively: ========================================= 1. dark frame 2. dark frame 3. flash 4. dark frame Same as before? Not exactly! Here is how the frame with flash will look like: ################################### ################################### ################################### ################################### ################################### ################################### ################################### ################################### ################################### ################################### ################################### ################################### Every single line will be lit by flash (the short burst of light also freezes the movement, so the motion blur would not be visible at all). Progressive frame captured, recorded interlaced as in case with HD7: ========================================================== 1. dark frame 2. flash 3. flash 4. dark frame Why? Because the progressive frame was divided into two half frames and recorded into two interlaced frames on whatever media it is. Here is what you see in the first flash-lit frame (note odd lines are lit): ################################### ...........dark line................. ################################### ...........dark line................. ################################### ...........dark line................. ################################### ...........dark line................. ################################### ...........dark line................. ################################### ...........dark line................. ################################### and the second one (note even lines are lit): ...........dark line................. ################################### ...........dark line................. ################################### ...........dark line................. ################################### ...........dark line................. ################################### ...........dark line................. ################################### ...........dark line................. ################################### And that is exactly what I have seen with HD7. Two consecutive frames lit by flash! So it is clearly a progressive imager recording progressively captured frames as interlaced video. Just as I hoped for. I went one step further to absolutely prove my point - I combined manually the flash-lit even and odd lines from the two frames into one and it gave me a perfect progressive frame without any comb effect on moving edges! Hurray! Here is a link to a series of 4 bitmaps that demonstrate what I just said, these are uncompressed captured frames. See how the lit lines in frame 2 and 3 stay in the same exact location on screen even though the dark lines move? That means that the odd lines are encoded into first interlaced frame and the even ones are encoded into the second frame. http://www.transferbigfiles.com/Down...b-a92f26a8c6fc Sorry for the long post, but I thought this could be useful for many of us who had no means of proving their suspicions... Have a nice day, Tibor Duliskovich |
|
|
|
|
April 14th, 2007, 02:44 PM
|
#188 |
|
Major Player
Join Date: Jan 2004
Location: Seattle
Posts: 243
|
Tibor
You sir are a genius. Thank you and patent that idea. |
|
|
|
|
April 14th, 2007, 07:04 PM
|
#189 |
|
New Boot
Join Date: Mar 2007
Location: Jacksonville, Florida, USA
Posts: 24
|
More samples
I posted several additional sample clips for HD7 here:
http://www.transferbigfiles.com/Down...3-18badb5ecee2 Grasshoppers on a flower, these are real small ones, about 10mm long, the lens cap was touching the flower, tele-macro enabled, gain manually +2 (it is a little too much, but in strong sunlight I did not see the LCD and thought that the black grasshoppers are under exposed). Grasshopper, sun was behind the clouds, a little soft: http://www.transferbigfiles.com/Down...3-9d42a09fae10 Owl, wind blowing the leaves, visible blocking and coding artifacts: http://www.transferbigfiles.com/Down...9-393288bba3bd |
|
|
|
|
April 14th, 2007, 07:06 PM
|
#190 |
|
New Boot
Join Date: Mar 2007
Location: Jacksonville, Florida, USA
Posts: 24
|
Editing Cyberlink and Vegas and TOD files
Here is my experience so far with editing SW. The bundled Cyberlink Producer was able to open TOD and TPD files. When you add them to a project Cyberlink also converts these files to regular MPG files. It does not recompress (I know because it happens too fast for my old Dell XPS 5), it simply puts the video into a new container. The file size will be almost exactly the same too.
Why is that important? Because Vegas 7d cannot directly open TOD files, says "unsupported file format", I tried renaming the TOD into MPG, Vegas could not open that either. The MPG from Cyberlinks cache folder Vegas opened without problems. Unfortunately Vegas and Cyberlink both open the file incorrectly, as an interlaced file. I tried creating progressive project and non-progressive, upper or lower field first, I tried in each of those projects mark the media properties, upper, lower or progressive and rendered as uncompressed AVI, but the result was always interlaced. So while these apps can open the TOD files, they do not interpret them correctly and there is no way manually to correct that. I spent about three hours doing all these tests, so I am confident. I do not have Premiere or any other editing SW so am curious if someone has any experience with them. |
|
|
|
|
April 15th, 2007, 01:40 PM
|
#191 |
|
Trustee
Join Date: Dec 2005
Location: Massachusetts
Posts: 1,773
|
I have never used this camcorder yet but I bet I can easily produce better footage than most people can with their HV20 or their HC7 but that’s if it’s on a tripod. Off a tripod, I don’t know. From all the first hand accounts I’ve read, something must be wrong with the stabilizer. Either that or the camcorder must be hard to hold for most people. I really think JVC needs to give an explanation for this.
__________________
https://www.youtube.com/user/PhotoVi...esEtc/featured https://www.pond5.com/artist/paulot Last edited by Paulo Teixeira; April 15th, 2007 at 03:35 PM. |
|
|
|
|
April 15th, 2007, 05:28 PM
|
#192 |
|
Trustee
Join Date: Jan 2002
Location: Pembroke Pines, Florida
Posts: 1,418
|
How does a Mac user view those TOD files?
|
|
|
|
|
April 15th, 2007, 05:53 PM
|
#193 |
|
Trustee
Join Date: Jan 2002
Location: Pembroke Pines, Florida
Posts: 1,418
|
Ignore previous post- VLC player plays them!
I'm seeing blown out highlights as noted by others. Colors and resolution is nice as long as within camera exposure limits- otherwise blown out! Reminds me of an HD10U but in 1080i and to internal HD- not quite the quality I expected or was hoping for. Having said that- in the right lighting and studio setup I'm sure the camera will deliver excellent results. The HV10 & HV20 have this camera beat as well as the FX1 & FX7. Not a bad cam at sub $1600 prices but not really ready for the masses and semi-pros (prosumers)......we'll have to wait for Canon, Panasonic or Sony to really make a winner in the hi-def internal HD cam cataegory! If by early summer no one else has anything comparable I'll get one- but I'll sell it the second Canon releases their version of an internal HD cam. |
|
|
|
|
April 15th, 2007, 07:15 PM
|
#194 |
|
New Boot
Join Date: Mar 2007
Location: Jacksonville, Florida, USA
Posts: 24
|
More observations and samples
Auto focus, bicycle approaching quite fast, zoom x10, tripod, focus is doing good job:
http://www.transferbigfiles.com/Down...c-3d6157f2bc05 Some quite low light footage, after sunset, heavy clouds, 10x zoom, tripod, river, boat: http://www.transferbigfiles.com/Down...e-478893ab6d33 Same scene at 6x zoom: http://www.transferbigfiles.com/Down...1-61bdf81ab55b Flower, macro, in a shade: http://www.transferbigfiles.com/Down...0-7e0963d065dc I noticed that at wide open aperture the lens is not stellar: it becomes very-very soft in corners. Add to this poor performance of OIS and I can only say, that Fujinon "professional" lens is the weakest link in HD7. I would only use it at F5.6 and F8.0 (it does not go beyond that either...). Regarding digital zoom. The sensor has given number of pixels it can resolve, so technically speaking digital zoom cannot add anything to image content, it simply blows up pixels (interpolates them). But there is a catch and it is in recording format. There is a given bandwidth to record the content onto media - it is a 30 mbps VBR limit. Consider these two scenatious: 1. capturing all pixels (representing lots of detail in optical image), encoding full frame with lots of details into 30 mbps stream and then, after the fact, enlarging part of frame, 2. capturing part of pixels (representing less optical detail) and encoding them with full bandwidth. It is obvious that in second case those fewer pixel can be encoded with higher quality, then the larger number of pixels in first scenario. So that is why digital zoom makes sense to certain degree. I estimate based on my tests with HD7, that an additional 2x zoom, extending the originaly 10x optical zoom into 20x - that is the max that makes sense. For your own consideration: 10x optical zoom, owl: http://www.transferbigfiles.com/Down...1-2ef3f4d6bfd6 20x time optical zoom, same owl, same distance: http://www.transferbigfiles.com/Down...c-039ebfea8071 Believe me, going beyond that did not improve the image at all. |
|
|
|
|
April 15th, 2007, 08:14 PM
|
#195 |
|
Trustee
Join Date: Jan 2002
Location: Pembroke Pines, Florida
Posts: 1,418
|
Tibor, your clips look good because the lighting was even and not too contrasty. It's the direct sunlight on flowers and objects as well as brightly lit objects that draw the comments.
Thanks for those clips- it does show potential for the JVC with subdued lighting. |
|
|
|
 |
| ||||||
|
||||||
|
|
