Go Back   Steve's Digicams Forums > Digital Cameras (Point and Shoot) > Sanyo

Reply
 
Thread Tools Search this Thread
Old Sep 28, 2007, 5:10 PM   #1
Senior Member
 
Join Date: Aug 2007
Posts: 101
Default

http://www.watch.impress.co.jp/av/do...6/zooma324.htm

anyone reads Japanese?

from what i can understand, the stills that are 16:9 are stills from the movie files... the standard 4:3 are the actual still photos from photo mode. There's also one or two interpolated images that looks crappy at 100%, but overall the photos seems less grainy than my E1/CA65

* the movie files on the page looks like its been transcoded to a lower quality... other samples i've seen so far looks far better than what impress has posted.
ray301 is offline   Reply With Quote
Sponsored Links
Old Sep 29, 2007, 11:19 AM   #2
Senior Member
 
Wayne12's Avatar
 
Join Date: Jun 2006
Posts: 1,071
Default

Here is an translation:

http://babelfish.altavista.com/tr?tt...2Fzooma324.htm

Can anybody tell us what it is about, I have given up on google translation, I seem to have more and more problems reading them, instead of getting better, but I still have trouble with this fish.

The pictures:

This one, looks through the leaves to the sky, there is that broad blue halo I noticed in the French show footage. It seems to be in many pictures and varies with brightness. I guess the lens.

But the latitude and colourisation finally appears to be there, but certain shots in video appear to be washed out an bit, like the tele sample:

http://www.watch.impress.co.jp/av/do...926/ezsp02.jpg

And this one:
http://watch.impress.co.jp/av/docs/20070926/ezsp13.jpg

Why?

The latitude is good (though there is not much light to judge properly) you can see under the trees and people walking under them, and the light on the trees, and umbrella with things in range.

Here we see clear burnout under the cat, but very good compared to previouse, but wonder how good.

http://www.watch.impress.co.jp/av/do...926/ezsp14.jpg

If it was not for this burnout, I would wonder if they had not used the new Altasens chip (I think there maybe an 4MP version, but am unsure) but other companies can produce such images I feel.

http://www.watch.impress.co.jp/av/do...926/ezsp11.jpg

Here, look around the girls nose, you see blue, look at the background, you see the disconcerting patterning across the field, but in the flowers the blur halo around them.

http://www.watch.impress.co.jp/av/do...926/ezsp15.jpg

Not truly professional video level of latitude (see the background) but beautiful warm filmic image, beautiful .

They seems to speak of many artifact problems and even noise problems in dark places (which I thought would not be the case given the setup, and if using altasens sensor, but I do not know the performance of the consumer sensor line). I can't verify this, will need much more in depth study of images and footage.

Preliminary Verdict, on these pictures only (not on handling etc). Consumer camera, good as consumer camera (despite lens and any small artifacts, yet to be examined). But as professional camera, that blue halo, the odd halo around the blurred flowers, and spotted sort of back ground field in compression, any artifact problem, and not so much latitude this time, an uncompressed HDMI output would save it (except the blue halo). I cannot follow what they are meaning about compression, but there are different levels of H264, each with different quality levels, and it depends on the implementation at whatever level is used. It is further claimed ambarella uses nice H264 compression, but I think FullHD 1080i, and maybe 720p60, 12mb/s is insufficient. 18mb/s in much more preferable, and at this stage I would say 24mb/s, with truly professional being 36mb/s+.

All in all, very good effort, but that blue halo I don't like.

------

By the looks of it, some of the clips were transcoded to HDV because of problems with playing them back.

------

Now the most important question for video, uncompressed HDMI out, maybe there will be h264 intra/cineform hardware capture solutions one day.
Wayne12 is offline   Reply With Quote
Old Sep 29, 2007, 12:07 PM   #3
Senior Member
 
Wayne12's Avatar
 
Join Date: Jun 2006
Posts: 1,071
Default

Forgot, most of those beautiful warm pictures are stills, I wish video used color like that. As we can remember, the still image color and latitude of the earlier HD cameras looked significantly better than video. Am still downloading the footage to find out what they video is like on this camera.
Wayne12 is offline   Reply With Quote
Old Sep 30, 2007, 9:23 AM   #4
Senior Member
 
Wayne12's Avatar
 
Join Date: Jun 2006
Posts: 1,071
Default

I have examined the, water torture, video test samples. That did that make the codec trip up. I tried VLC to image function, but the codec corruption turns up in the PNG still stream, but found the same underlying disjointed blocking in the videos using Power DVD (works good, don't know which loaded codec it is tapping into).

That scene is definitely consumer quality, in no way professional. The consumer quality here, is not what I would call high. It is an extreme test, though should not be, and I hope this sort of thing doesn't turn up in other normal scenes. Much of the movement of the water shapes have an predictable flow, it is not the surf, or the surface of an choppy lake.

From the calculations, the compression ratio of the VGAp60 scene is roughly the same as the 720p30 mode, but should be an bit more efficient because it is 60fps. The compression of the 720p60 scene is significantly more. This would make me think that even if it had 18mb/s in the 720p60 mode it would have similar results. As an generalisation, looks like the way to guarantee results as good as another codec, is to use the same data rate.

Have yet to look at the HDV transcoded clips, and will not be looking at the 1080i clips, but I don't expect them to be any better.

Intermediary results, consumer, even more consumer under stress (it is the soccer/sports filming I am worried about). But great image colourisation and latitude possibilities, depending on the remaining clips. Codec consumer, like an low end AVCHD camera. Image like an good consumer camera.

Sorry for the extensive analysis, but I have been waiting for an very long time to analysis the footage from this camera. This camera is still an big step in the right direction for Sanyo.
Wayne12 is offline   Reply With Quote
Old Oct 1, 2007, 4:50 AM   #5
Senior Member
 
Join Date: Aug 2007
Posts: 101
Default

Wayne12 wrote:
Quote:
I have examined the, water torture, video test samples. That did that make the codec trip up. I tried VLC to image function, but the codec corruption turns up in the PNG still stream, but found the same underlying disjointed blocking in the videos using Power DVD (works good, don't know which loaded codec it is tapping into).

That scene is definitely consumer quality, in no way professional. The consumer quality here, is not what I would call high. It is an extreme test, though should not be, and I hope this sort of thing doesn't turn up in other normal scenes. Much of the movement of the water shapes have an predictable flow, it is not the surf, or the surface of an choppy lake.

From the calculations, the compression ratio of the VGAp60 scene is roughly the same as the 720p30 mode, but should be an bit more efficient because it is 60fps. The compression of the 720p60 scene is significantly more. This would make me think that even if it had 18mb/s in the 720p60 mode it would have similar results. As an generalisation, looks like the way to guarantee results as good as another codec, is to use the same data rate.

Have yet to look at the HDV transcoded clips, and will not be looking at the 1080i clips, but I don't expect them to be any better.

Intermediary results, consumer, even more consumer under stress (it is the soccer/sports filming I am worried about). But great image colourisation and latitude possibilities, depending on the remaining clips. Codec consumer, like an low end AVCHD camera. Image like an good consumer camera.

Sorry for the extensive analysis, but I have been waiting for an very long time to analysis the footage from this camera. This camera is still an big step in the right direction for Sanyo.
Have you examined the 1080i sample from techvideoblog.com?
http://charbax.videohost.eu/ifa2007/SANY0003.MP4

and the 720p 60fps sample from digiblo.jp?
http://www.digiblo.jp/archives/mov/c.../sample720.zip

i saw these two samples first before i saw watch impress page, that's why i actually thought impress did something wacky with their sample clips. Hopefully more sample footage will be available soon so i can plop my money down on it
ray301 is offline   Reply With Quote
Old Oct 2, 2007, 1:56 AM   #6
Senior Member
 
Wayne12's Avatar
 
Join Date: Jun 2006
Posts: 1,071
Default

The impress videos were trans-coded to HDV for better playback, but are looking nice (still downloading that massive file). They are better test videos, I can tell an number of things from them, and any camera that can do those shots successfully is doing pretty good. The native format water video is something that low end codecs have an lot of trouble with, but with he amount of h264 data on hand, I didn't expect it to fail so much. Even with HDV you will probably see some stepping/blocking (because many HDV cameras don't handle predictable transforming water shapes that well, but h264 is many years newer than Mpeg2 technology). I wish that companies would preserve the shape (and edge) of features at the expense of details, rather than let images break down to blocking, it looks better.

I did view the other clips first, but because of problems on my machine it was just choppy (now runs an bit less choppy) I saw very little of the sanyo003, as it was too large and mostly unplayable, and an interview (I was more interested in what was said. The 720p60 clip is not such an good test video, the color and much of it looks murky (because of the content) there is precious little movement, and apart from the water warping, little is obvious.

I am yet to examine all the stills from the main test (to see size 1080i rather than 960*540 on my monitor) but even in HDV it looks pleasantly good compared to the native water test. The only disappointment at the moment is the blue halo (I assume the saturation level can be turned down) as what was in the water test is rarely seen so far. So, this might be turning into an good consumer camera for you. I do notice some crawling at 960*540, I still see something that looks like fly-screen type crawling, but that could be my systems downscaling, so I have to view the PNG lossless stills from VLC. I am examining the thing from for upper end consumer quality, and lower end professional quality, leaving out lots of the stuff that an upper professional review would pick on. As long as it looks strikingly stunning, with little, or no obvious artifacting, or other problems, it's good, it doesn't need to be perfect.
Wayne12 is offline   Reply With Quote
Old Oct 2, 2007, 2:20 AM   #7
Senior Member
 
Wayne12's Avatar
 
Join Date: Jun 2006
Posts: 1,071
Default

Posting and checking those links has managed to reset my download of the main test again (was nearly finished). I might post the finale assessment without going through the entire test.
Wayne12 is offline   Reply With Quote
Old Oct 2, 2007, 2:36 AM   #8
Senior Member
 
Join Date: Aug 2007
Posts: 101
Default

Wayne12 wrote:
Quote:
Posting and checking those links has managed to reset my download of the main test again (was nearly finished). I might post the finale assessment without going through the entire test.
I did notice one thing with the water test footage when i played it back on media player classic... Once the file plays to the end and stops, the image that it pauses on is flawless, the water shapes are correct and no noise artifacts exists. Could it be a decoder issue?
ray301 is offline   Reply With Quote
Old Oct 2, 2007, 8:09 AM   #9
Senior Member
 
Wayne12's Avatar
 
Join Date: Jun 2006
Posts: 1,071
Default

I would like to examine that again. It could be it has ended on an GOP start frame (is that the term) which is the highest quality frame, that is intra compressed, the other frames are lower quality and is where most of the problems occur. Those frames are inter frames that mainly use difference compression that only handle limited difference (movement etc) and are based on the difference from the intra frames (which is why pure intra frame compression codecs work better, but use an lot more data). It is an little bit more complicated than that, there are multiple sub-types of frames as well. The best compromise is an variable compression codec that keeps down size but maintains compression quality and does intra frames as often as needed.
Wayne12 is offline   Reply With Quote
Old Oct 2, 2007, 12:23 PM   #10
Junior Member
 
Join Date: Sep 2007
Posts: 21
Default

Quote:
I did notice one thing with the water test footage when i played it back on media player classic... Once the file plays to the end and stops, the image that it pauses on is flawless, the water shapes are correct and no noise artifacts exists. Could it be a decoder issue?
same problems here when playing. im using macbook.
T4R06 is offline   Reply With Quote
 
Reply


Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off



All times are GMT -5. The time now is 2:28 PM.