Sony NEX-VG10 OFFICIAL Camcorder Thread - Page 7 - AVS Forum
Forum Jump: 
 
Thread Tools
Old 09-12-2010, 01:54 PM
Senior Member
 
Jay968's Avatar
 
Join Date: Jun 2010
Posts: 294
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 12
Something I really don't understand...everything I read about this camera suggests that the way it handles 30p and wraps into a 60i container means that there is no interlace artifacting. However, everything I have seen from this camera so far shows so much interlace artifacting that it appears almost unuseable!! Could this be due to the final statement in the following (after the "Another quick note...)?

>>Besides the interchangeable lens, the NEX-VG10 features all the typical things you’d expect in a compact handycam. Proper on-camera mic, proper audio headphone monitoring and external mic inputs, an LCD and button layout that videographers will understand, and of course a form factor that lends itself to handheld shooting. Storage is on SD and MemoryStick media via a single slot, and the video codec is AVCHD at 24Mbps. It is a 1920x1080 full-raster image, as you might expect. Modes include 1920x1080 @ 60i, and bitrates offered are FX @24Mbps and FH @17Mbps. There is also a 1440x1080 @60i mode at 9Mbps. No word on 24p, unfortunately. It seems Sony is leaving it out of this model. But you should note that the 60i material originates from a progressive sensor, and images do not appear to show any tearing or interlace artifacting. This is similar to how the NEX-5 works…it’s video mode offers 60i as well, but it looks progressive on your NLE timeline. Another quick note: I also have it on good authority from a Sony contact that the mini-HDMI port output is pre-compression, so you also have the option of uncompressed recording when tethered to a card or capture device.<<

From http://provideocoalition.com/index.p...g10_announced/
Jay968 is offline  
Sponsored Links
Advertisement
 
Old 09-12-2010, 02:40 PM
AVS Addicted Member
 
Ken Ross's Avatar
 
Join Date: Nov 2000
Location: N.Y.
Posts: 24,639
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 1419 Post(s)
Liked: 1534
Quote:
Originally Posted by Jay968 View Post

I did leave a little bit unsaid when I wrote what I did yesterday. I put a card inside the camera and held onto it till I could bring it home and look at the footage closer after putting it onto a disk and then viewing on a larger TV screen. The footage is indeed reminding me of what I would expect to get out of a Canon 5DMKII (which I own and haven't used as a video camera since I played around with it as such several months ago and ultimately rejected). I find the 30p footage to be absolutely horrendous and to this day just cannot get into my thick head why anyone would actually choose to shoot video in that manner when technology has gotten us to the point where we can instead shoot in such gorgeous HD video otherwise. It's like 'lets make our video look like decades old film stock instead of taking advantage of the latest technology which makes HD video look like we are looking through a window at the world.' I just don't get it. To each his own I spose.

Couldn't agree more. This is why I caution anyone to watch this footage on a big screen HDTV to see what it looks like there. It could be a shocker as you discovered. This entire 24p/30p thing has always eluded me.

Additionally, Paulo's link shows just how bad autofocus is in the cam (poor camera work aside). As I said before, it's not just the Sony, it's any DSLR cam too. I just don't see the interchangeable lens capability offsetting these major issues IMO.

Considering the artifacts I see in my NEX5 videos, I'd also be surprised if the VG10 is free of them.
Ken Ross is offline  
Old 09-12-2010, 03:32 PM
Senior Member
 
Jay968's Avatar
 
Join Date: Jun 2010
Posts: 294
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 12
The thing that is puzzling me the most though is the fact that the way this camera handles 30p, by wrapping it into a 60i container is supposed to mean that since it is essentially 30p there should be no aliasing artifacts. I am finding anything BUT this to be the case. As a matter of fact the jaggies that I am seeing are horrendous to say the least, among the worst I have ever seen in a video camera.
Jay968 is offline  
Old 09-13-2010, 11:19 AM
AVS Addicted Member
 
Ken Ross's Avatar
 
Join Date: Nov 2000
Location: N.Y.
Posts: 24,639
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 1419 Post(s)
Liked: 1534
Jay, I don't get it either. If the artifacts are anything like I've seen on my NEX5, they are pretty bad.
Ken Ross is offline  
Old 09-13-2010, 11:28 AM
AVS Special Member
 
Ungermann's Avatar
 
Join Date: Aug 2006
Location: San Diego, California
Posts: 3,956
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 118 Post(s)
Liked: 29
I have no problem with 30p. It is more fluid than 24p, requires less care when setting the shot, and it is universally supported all over the Web. If someone still thinks that shooting for the Web is a niche market he has to wake up. Shooting for broadcast as we knew it will soon become a niche market.
Ungermann is offline  
Old 09-13-2010, 01:36 PM
AVS Addicted Member
 
Ken Ross's Avatar
 
Join Date: Nov 2000
Location: N.Y.
Posts: 24,639
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 1419 Post(s)
Liked: 1534
There is no way that you can fluid playback with any kind of rapid motion or panning like that of 60i or 60p with either 24p or 30p (even though 30p is a bit better than 24p) on a large screen HDTV. Yes, it's better on a computer monitor, but I've yet to see 30p look smooth on any large HDTV I've ever seen.

These formats can not be used for smooth motion with any consistency for 'run & gun' shooting like many users will do.
Ken Ross is offline  
Old 09-13-2010, 01:41 PM
AVS Special Member
 
Paulo Teixeira's Avatar
 
Join Date: May 2007
Location: Massachusetts
Posts: 2,728
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 31 Post(s)
Liked: 33
I also posted a link to the Watch.Impress review which has sample as well.

As far as the recent link I posted to, to be fair, that's a 25p clip and the US cameras will obviously get 30p so the motion could have been better.
Paulo Teixeira is offline  
Old 09-13-2010, 02:35 PM
AVS Special Member
 
Ungermann's Avatar
 
Join Date: Aug 2006
Location: San Diego, California
Posts: 3,956
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 118 Post(s)
Liked: 29
Quote:
Originally Posted by Ken Ross View Post

There is no way that you can fluid playback with any kind of rapid motion or panning like that of 60i or 60p with either 24p or 30p (even though 30p is a bit better than 24p) on a large screen HDTV. Yes, it's better on a computer monitor, but I've yet to see 30p look smooth on any large HDTV I've ever seen.

These formats can not be used for smooth motion with any consistency for 'run & gun' shooting like many users will do.

As I said, you -- either willingly or unwillingly -- do not seem to see strobing in "proper" movies. Obviously, good framing and camera movement help a lot, keeping your attention on a main character, not on background. The VG10 offers shallow DOF similar to proper movie cameras, which helps to mitigate strobing. I watch quite a lot of 24p, 25p and 30p on my 50-inch TV. There are better videos and worse ones, but if done properly, 30p looks just fine. But limiting users to just one frame rate is nonsensical in 2010. Shame on Sony.
Ungermann is offline  
Old 09-13-2010, 03:33 PM
AVS Addicted Member
 
Ken Ross's Avatar
 
Join Date: Nov 2000
Location: N.Y.
Posts: 24,639
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 1419 Post(s)
Liked: 1534
Quote:
Originally Posted by Ungermann View Post

As I said, you -- either willingly or unwillingly -- do not seem to see strobing in "proper" movies. Obviously, good framing and camera movement help a lot, keeping your attention on a main character, not on background. The VG10 offers shallow DOF similar to proper movie cameras, which helps to mitigate strobing. I watch quite a lot of 24p, 25p and 30p on my 50-inch TV. There are better videos and worse ones, but if done properly, 30p looks just fine. But limiting users to just one frame rate is nonsensical in 2010. Shame on Sony.

My main point is that 30p can not handle rapid movement as smoothly as 60i or 60p. It's simply not technically possible with a restricted frame rate. There is no need for 'careful scene setup' to achieve smooth motion with either 60i or 60p. Anyone that does 'run & gun' will know what I'm talking about. When shooting parties, sports, kids etc., you don't have the luxury of 'careful scene setup' and rapid motion will likely suffer. For people not making 'movies' and simply shooting on the fly, this can be a real issue.

I do agree that it's nuts that Sony didn't provide for alternate frame rates.
Ken Ross is offline  
Old 09-13-2010, 03:40 PM
AVS Special Member
 
Ungermann's Avatar
 
Join Date: Aug 2006
Location: San Diego, California
Posts: 3,956
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 118 Post(s)
Liked: 29
Quote:
Originally Posted by Ken Ross View Post

There is no need for 'careful scene setup' to achieve smooth motion with either 60i or 60p. Anyone that does 'run & gun' will know what I'm talking about. When shooting parties, sports, kids etc., you don't have the luxury of 'careful scene setup' and rapid motion will likely suffer.

Then one puts his shakycam stuff on YouTube and it strobes like hell. Wait a minute, you don't put your stuff on YouTube. I guess there is a reason for that ;-)
Ungermann is offline  
Old 09-14-2010, 08:46 PM - Thread Starter
Senior Member
 
hpmoon's Avatar
 
Join Date: Dec 2006
Location: Washington, D.C.
Posts: 254
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 10
Quote:
Originally Posted by Ungermann View Post

In 2006, when AVCHD was introduced...etc....etc....

Ungermann, your posting is eerily aggressive, and I really haven't been following them too closely (since my interest is in, oh, well, the topic of this thread), but basically what it seems to come down to is this: AVCHD imposes significant demands on CPUs/GPUs during editing while offering a high rate of compression with decent performance; but it's not the only game in town. As an aside, D-SLRs use a ****** codec that performs worse than AVCHD, even at higher bitrates. (That debate is so old it smells like a fresh fart, so please for the love of god don't strike a match and ignite it into a flamethrower.) Finally, there are codecs marginally superior to AVCHD that require more disk space but smaller processing demands.

What you need to factor out is that the current state-of-the-art, de facto, has made the CPU/GPU demands of AVCHD moot. For anyone who finally leaves behind that Apple fanboy dinosaur Final Cut Pro to the graveyard and seizes the power of Adobe Premiere Pro CS5 with its Mercury Playback Engine paired with GPU acceleration, AVCHD footage flies through a timeline like a breeze.

That you have an outdated computer, and/or brand loyalty to Steve Jobs, has nothing to do with AVCHD.

The trivial cost of upgrading computer hardware, compared to the price gouging behaviors of these video equipment manufacturers, justifies offloading demands to the post-production phase. As such, I don't include lens filters and on-location color concerns anymore; it's all to be done in post, using even a plain vanilla Core i7 system via Adobe products and GPU acceleration sealing the deal.

Time to arrive at 2010.
hpmoon is offline  
Old 09-14-2010, 08:53 PM - Thread Starter
Senior Member
 
hpmoon's Avatar
 
Join Date: Dec 2006
Location: Washington, D.C.
Posts: 254
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 10
Quote:
Originally Posted by Jay968 View Post

The thing that is puzzling me the most though is the fact that the way this camera handles 30p, by wrapping it into a 60i container is supposed to mean that since it is essentially 30p there should be no aliasing artifacts. I am finding anything BUT this to be the case. As a matter of fact the jaggies that I am seeing are horrendous to say the least, among the worst I have ever seen in a video camera.

Jay968, you are confusing two unrelated phenomena. In another post, you wrote that you were seeing interlacing artifacts all over the place. Then here you are using the term aliasing while going back to the 30p/60i conundrum. Back to the drawing boards for you.

Anyway, I haven't seen a single example (albeit in the compressed streaming footage online) of any aliasing OR interlacing distortions.
hpmoon is offline  
Old 09-14-2010, 08:56 PM - Thread Starter
Senior Member
 
hpmoon's Avatar
 
Join Date: Dec 2006
Location: Washington, D.C.
Posts: 254
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 10
I've updated the original post with the English-language Operation Guide and Handbook in PDF format. The Operation Guide comes printed with the NEX-VG10, while the Handbook comes on disk as a PDF.

These things are shipping out now (domestic U.S.); I got my FedEx tracking number an hour ago.

All the sample videos so far are collected at: http://www.nexvg10.info.
hpmoon is offline  
Old 09-14-2010, 09:50 PM
Senior Member
 
Jay968's Avatar
 
Join Date: Jun 2010
Posts: 294
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 12
Quote:
Originally Posted by hpmoon View Post

Jay968, you are confusing two unrelated phenomena. In another post, you wrote that you were seeing interlacing artifacts all over the place. Then here you are using the term aliasing while going back to the 30p/60i conundrum. Back to the drawing boards for you.

Anyway, I haven't seen a single example (albeit in the compressed streaming footage online) of any aliasing OR interlacing distortions.

Excuse me for using improper terms, I am not as well versed as I wish I was.

According to Wikepedia --- "Jaggies" is the informal name for artifacts in raster images, most frequently from aliasing,[1] which in turn is often caused by non-linear mixing effects producing high-frequency components and/or missing or poor anti-aliasing filtering prior to sampling.

Anyway, to make a long story short, I have now shot with the camcorder for 2 days and what I am seeing in excess are jaggies on telephone cables, white lines in the streets, building edges etc. much worse than any camcorder I have ever seen. The best way to describe it is to compare its look to that of a standard definition tv show being displayed on a high definition 55" LCD TV. When I view such programming, I see jaggies all over the place. Same here with this Sony. I own a Canon 5D MKII and do not use that camera as a camcorder for exactly this reason. It does the jaggie deal as well and is known for that. I did some footage yesterday comparing the two side by side. The Sony was a bit worse. On the other hand, I see a lot of this on the news these days when they go on location. It appears as if they are using cheaper, more consumer oriented cameras due to budget constraints. What a shame.

I HAVE read that the way the Sony shoots 30p and wraps into a 60i container is supposed to eliminate such artifacts since 30p is progressive and it is then viewed on a progressive LCD TV set. I would also assume that since it's a camcorder it would have more aggressive filtration than a 5DMKII. If I am wrong in these assumptions, I apologize in advance.

You say you haven't seen any interlacing or aliasing distortions. Have you seen anything on a large screen HDTV? If all you have seen are online samples, with all due respect, you have not seen what this camcorder looks like. I am sorry but viewing footage only online gives no indication of what a camcorder can or cannot do.

For what it's worth, there is another unrelated problem that I have been seeing with the camcorder that I have had the opportunity to borrow...every once in a while, more times than I care to report, the buttons just freeze up. I can press the menu button, nothing happens, press a different button, nothing happens, wait a few minutes and press one of them again and all is well. Several minutes later, some button, maybe one of the above, maybe another...again, nothing. One time the start/stop button wouldn't even let me stop the video I was recording. None of the other buttons did anything either. I had to turn the camcorder off and then back on again to gain back control of it. So either this camera that I have has some issues, or there is a software problem with them. Time will tell. I suppose too that since I do have the button problem AND the jaggie problem that perhaps I am playing with a defective camcorder...or perhaps there ARE software issues related to both problems. Or, perhaps since this problem seems to 'usually' occur shortly after I have recorded something, maybe the camcorder doesn't allow you to do the buttons until it is done writing the footage to the card (?). Resetting the camcorder to its defaults did nothing to correct anything btw.
Jay968 is offline  
Old 09-14-2010, 09:56 PM - Thread Starter
Senior Member
 
hpmoon's Avatar
 
Join Date: Dec 2006
Location: Washington, D.C.
Posts: 254
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 10
Quote:
Originally Posted by Jay968 View Post

Excuse me for using improper terms, I am not as well versed as I wish I was, but to make a long story short, I have now shot with the camcorder for 2 days and what I am seeing in excess are jaggies on telephone cables, white lines in the streets, building edges etc. much worse than any camcorder I have ever seen. The best way to describe it is to compare its look to that of a standard definition tv show being displayed on a high definition 55" LCD TV.

I HAVE read that the way it shoots 30p and wraps into a 60i container is supposed to eliminate such artifacts since 30p is progressive and it is then viewed on a progressive LCD TV set. If I am wrong in this assumption, I apologize in advance.

At any rate...

According to Wikepedia --- "Jaggies" is the informal name for artifacts in raster images, most frequently from aliasing,[1] which in turn is often caused by non-linear mixing effects producing high-frequency components and/or missing or poor anti-aliasing filtering prior to sampling. "Jaggies" also refer to a particular kind of screen recognition error on touch-sensitive PDAs and Smartphones.[2

You say you haven't seen any interlacing or aliasing distortions. Have you worked with the camcorder? Have you viewed any footage on a large screen HDTV? If all you have seen are online samples, with all due respect, you have not seen what this camcorder looks like. I am sorry but viewing footage only online gives no indication of what a camcorder can or cannot do.

These issues are of course highly technical. Then again, this forum is called AV Science, so...

The only proper way to judge the capabilities are to put the following into your workflow as bare minimum standards without a single missing link:

1. AVCHD 1920 x 1080 @ 30fps and 24 Mbps;
2. Generous lighting and atmospheric conditions so that the sensor is not straining;
3. Either:
a. Direct output via HDMI to an HDTV's HDMI input; or
b. Rendering from a properly configured timeline (matching 100% the specs of the source) to a progressive AVC H.264 format with at least a 10 Mbps bitrate.

Without that, test results are meaningless.

And of course, an assertion that the footage is poor requires comparison to what is believed superior.
hpmoon is offline  
Old 09-14-2010, 10:40 PM
Senior Member
 
Jay968's Avatar
 
Join Date: Jun 2010
Posts: 294
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 12
Yes I have actually done all those steps. I MAY have overlooked making sure that my timeline was rendered at progressive though! Not sure, I will have to double check that one. But on the other hand, direct output via HDMI to my HDMI TV input surely showed the jaggies...no question.

Doing some google searches on 'jaggies' in relation to these large sensor camcorders produces an awful lot of complaints about it in most all of the DSLR cameras, the 7D at 1080p60 being especially bad

BTW your quote (of my response) appears differently than my post due to my editing. I tried to be a bit more thorough in my response to you and to also give more info and insight into the camcorder and what I have seen with it.
Jay968 is offline  
Old 09-15-2010, 08:32 AM
AVS Addicted Member
 
amirm's Avatar
 
Join Date: Jan 2002
Location: Washington State
Posts: 18,375
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 981 Post(s)
Liked: 547
Jay, I did not see any interlace artifacts in the sample provided but I I did see the problem you now state with some diagonal lines broken (i.e. sampling/resampling error of some kind). It is hard though to separate what the hosting site does and the original content.

Are you able to post still images of original footage shot with this camera where the problem occurs?

Amir
Retired Technology Insider
Founder, Madrona Digital
"Insist on Quality Engineering"
amirm is offline  
Old 09-15-2010, 08:35 AM
AVS Addicted Member
 
amirm's Avatar
 
Join Date: Jan 2002
Location: Washington State
Posts: 18,375
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 981 Post(s)
Liked: 547
Quote:
Originally Posted by hpmoon View Post

The only proper way to judge the capabilities are to put the following into your workflow as bare minimum standards without a single missing link:

1. AVCHD 1920 x 1080 @ 30fps and 24 Mbps;
2. Generous lighting and atmospheric conditions so that the sensor is not straining;
3. Either:
a. Direct output via HDMI to an HDTV's HDMI input; or
b. Rendering from a properly configured timeline (matching 100% the specs of the source) to a progressive AVC H.264 format with at least a 10 Mbps bitrate.

Without that, test results are meaningless.

Well, they are also meaningless with step 3(b) just the same . We want to evaluate the original, not the recompressed version. The latter can introduce a lot of problems especially at the low data rate you mention. It could also mask problem due to additional (loop) filtering.

Quote:


And of course, an assertion that the footage is poor requires comparison to what is believed superior.

This is not necessary. We do not expect to see the types of problems he is talking about. Either there is a proper antialiasing stage or not.

Amir
Retired Technology Insider
Founder, Madrona Digital
"Insist on Quality Engineering"
amirm is offline  
Old 09-15-2010, 09:19 AM - Thread Starter
Senior Member
 
hpmoon's Avatar
 
Join Date: Dec 2006
Location: Washington, D.C.
Posts: 254
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 10
Quote:
Originally Posted by amirm View Post

Well, they are also meaningless with step 3(b) just the same . We want to evaluate the original, not the recompressed version. The latter can introduce a lot of problems especially at the low data rate you mention. It could also mask problem due to additional (loop) filtering.


This is not necessary. We do not expect to see the types of problems he is talking about. Either there is a proper antialiasing stage or not.

"Meaningless": You probably mean "less meaningful." I wouldn't be so sure that per the difference between, say, 9 Mbps (you call that a "low" bitrate) and 15 Mbps, you wouldn't fail a blind viewing test -- which leads to the other issue, that you say comparisons are not necessary. The idea is not that comparisons are necessary, rather that they're extremely useful in analyzing the relative merits of products that use this sensor technology (which is to say, every single video capture device in this class that is not immune to moire).
hpmoon is offline  
Old 09-15-2010, 10:15 AM
Senior Member
 
Jay968's Avatar
 
Join Date: Jun 2010
Posts: 294
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 12
Quote:
Originally Posted by amirm View Post

Jay, I did not see any interlace artifacts in the sample provided but I I did see the problem you now state with some diagonal lines broken (i.e. sampling/resampling error of some kind). It is hard though to separate what the hosting site does and the original content.

Are you able to post still images of original footage shot with this camera where the problem occurs?

Since I had only borrowed the camcorder for 2 days and I no longer have it, it's a bit difficult for me to do anything at this point. Let me take a look at what footage I had saved and see if I can post something someplace. I will say that playing with the sharpness, bringing it down a bit did help, but at the same time doing so left the footage looking less than ideal. It just may have to be the way to set this camera though.
Jay968 is offline  
Old 09-15-2010, 02:22 PM
AVS Addicted Member
 
amirm's Avatar
 
Join Date: Jan 2002
Location: Washington State
Posts: 18,375
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 981 Post(s)
Liked: 547
Quote:
Originally Posted by hpmoon View Post

"Meaningless": You probably mean "less meaningful." I wouldn't be so sure that per the difference between, say, 9 Mbps (you call that a "low" bitrate) and 15 Mbps, you wouldn't fail a blind viewing test -- which leads to the other issue, that you say comparisons are not necessary.

I think you may be confusing issues here. I am not debating quality difference but rather, what it takes to see the problem being mentioned. AVC employs a loop filter which especially at lower bitrates, filters high frequency detail. In that regard, it may very well soften the effect of aliasing distortion as to make it invisible.

What you are saying is that perceptually, the picture might look the same which might be true to some point, but definitely not in this regard where we want to know precisely what the camera is doing.

Quote:


The idea is not that comparisons are necessary, rather that they're extremely useful in analyzing the relative merits of products that use this sensor technology (which is to say, every single video capture device in this class that is not immune to moire).

A lot of things could cause such a problem. Let's figure it out first, before generalizing that it would also happen to other cameras with the same sensor. That analysis, can be done without comparison to other devices. You either get strairstep lines or you don't.

Amir
Retired Technology Insider
Founder, Madrona Digital
"Insist on Quality Engineering"
amirm is offline  
Old 09-15-2010, 02:26 PM - Thread Starter
Senior Member
 
hpmoon's Avatar
 
Join Date: Dec 2006
Location: Washington, D.C.
Posts: 254
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 10
Quote:
Originally Posted by amirm View Post

I think you may be confusing issues here. I am not debating quality difference but rather, what it takes to see the problem being mentioned. AVC employs a loop filter which especially at lower bitrates, filters high frequency detail. In that regard, it may very well soften the effect of aliasing distortion as to make it invisible.

What you are saying is that perceptually, the picture might look the same which might be true to some point, but definitely not in this regard where we want to know precisely what the camera is doing.


A lot of things could cause such a problem. Let's figure it out first, before generalizing that it would also happen to other cameras with the same sensor. That analysis, can be done without comparison to other devices. You either get strairstep lines or you don't.

I do agree with all you're saying in this regard.

Something that intrigues me -- a little off-topic -- is that next week's Photokina is rumored to be the launch of the NEX-7, which is widely reported as having a complete APS-C sensor re-design compared to the one in the NEX-3, NEX-5 and NEX-VG10 (which are allegedly all identical). Will the NEX-VG10 have been instantly outdated by this time next week?

Moreover, what are the "problems" being addressed in the re-design? Moire? Sensitivity?
hpmoon is offline  
Old 09-15-2010, 02:27 PM
AVS Addicted Member
 
amirm's Avatar
 
Join Date: Jan 2002
Location: Washington State
Posts: 18,375
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 981 Post(s)
Liked: 547
Quote:
Originally Posted by Jay968 View Post

Since I had only borrowed the camcorder for 2 days and I no longer have it, it's a bit difficult for me to do anything at this point. Let me take a look at what footage I had saved and see if I can post something someplace. I will say that playing with the sharpness, bringing it down a bit did help, but at the same time doing so left the footage looking less than ideal. It just may have to be the way to set this camera though.

You should not have to soften the image of course to get rid of it.

If you still have the footage, you can use editing software to capture a frame. Failing that, you can use print-screen type function but be sure to view the image 1:1 (i.e. full size/100%). And then compress the still losslessly (e.g. .png) and post it here. AVS will host the image for you as an attachment so you don't need a hosting site.

Amir
Retired Technology Insider
Founder, Madrona Digital
"Insist on Quality Engineering"
amirm is offline  
Old 09-15-2010, 02:42 PM
AVS Addicted Member
 
Ken Ross's Avatar
 
Join Date: Nov 2000
Location: N.Y.
Posts: 24,639
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 1419 Post(s)
Liked: 1534
Quote:
Originally Posted by hpmoon View Post

These issues are of course highly technical. Then again, this forum is called AV Science, so...

The only proper way to judge the capabilities are to put the following into your workflow as bare minimum standards without a single missing link:

1. AVCHD 1920 x 1080 @ 30fps and 24 Mbps;
2. Generous lighting and atmospheric conditions so that the sensor is not straining;
3. Either:
a. Direct output via HDMI to an HDTV's HDMI input; or
b. Rendering from a properly configured timeline (matching 100% the specs of the source) to a progressive AVC H.264 format with at least a 10 Mbps bitrate.

Without that, test results are meaningless.

And of course, an assertion that the footage is poor requires comparison to what is believed superior.

I have seen precisely the artifacts that Jay describes output from the NEX5 directly, via HDMI, to a 60" Pioneer Kuro plasma. Some of the worst artifacts I saw were when panning in bright outdoor light on a cruise ship. Many horizontal components of the ship suffered from the same 'jaggies' that Jay describes. As the camera pans, the artifacts take on an almost 'shimmering' effect...not pretty. This footage was not edited, not converted and simply displayed directly from the camera via HDMI. I also tried removing the memory card and displaying it on both a PS3 as well as a new Panasonic 3D BD player. All had the same issue.

This is certainly an issue with the NEX5's video and it appears to be similar with the VG10. It's hard for me to point a finger at just these Sony products as I've seen the same thing with Panasonic DSLRs, perhaps not quite as bad, when using their video capabilities.
Ken Ross is offline  
Old 09-15-2010, 02:49 PM - Thread Starter
Senior Member
 
hpmoon's Avatar
 
Join Date: Dec 2006
Location: Washington, D.C.
Posts: 254
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 10
Quote:
Originally Posted by Ken Ross View Post

I have seen precisely the artifacts that Jay describes output from the NEX5 directly, via HDMI, to a 60" Pioneer Kuro plasma. Some of the worst artifacts I saw were when panning in bright outdoor light on a cruise ship. Many horizontal components of the ship suffered from the same 'jaggies' that Jay describes. As the camera pans, the artifacts take on an almost 'shimmering' effect...not pretty. This footage was not edited, not converted and simply displayed directly from the camera via HDMI. I also tried removing the memory card and displaying it on both a PS3 as well as a new Panasonic 3D BD player. All had the same issue.

This is certainly an issue with the NEX5's video and it appears to be similar with the VG10. It's hard for me to point a finger at just these Sony products as I've seen the same thing with Panasonic DSLRs, perhaps not quite as bad, when using their video capabilities.

Another thing potentially to account for is that the built-in display engines of portable electronic devices are notoriously poor; even if dedicated microprocessors can manage incoming bitstreams, it is a whole other architecture for spitting back out content. The same goes for the signal pathway from flash memory through a consumer BD player/PS3 and into an HDTV's HDMI input.

The signal loss caused by laying footage into an NLE and rendering out at a high bitrate is so nominal that I'm sure it's preferable to the uncertainties introduced by judging footage via playback engines that are no match for Core i7 CPUs or, for that matter, a true mastered Blu-Ray disc.
hpmoon is offline  
Old 09-15-2010, 03:46 PM
Senior Member
 
Jay968's Avatar
 
Join Date: Jun 2010
Posts: 294
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 12
Frankly I think this is an issue with many consumer camcorders these days. Seems to have popped up since AVCHD came about. Some are worse than others. My Panasonic TM700 does it in all but 60p (even then there is SOME of it), I have seen it in the HF-S21 Canon as well. As I said earlier, it's one of the biggest negatives about using the 5DMKII in video mode as well. I have read articles and even seen videos with suggestions on how to deal with it, most saying throw the objects that you see the jaggies in out of focus, shoot with shallow depth-of-field etc.

Someone mentioned that you should not have to cut down on the sharpness to combat this. I couldn't agree more, you shouldn't. However reality is that this IS one way to help this situation. Remember a LOT of this has come about due to the resolution that the sensors are capable of these days. Putting a 14MP sensor in a video camera is asking for trouble.
Jay968 is offline  
Old 09-15-2010, 04:09 PM
AVS Addicted Member
 
Ken Ross's Avatar
 
Join Date: Nov 2000
Location: N.Y.
Posts: 24,639
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 1419 Post(s)
Liked: 1534
Quote:
Originally Posted by hpmoon View Post

Another thing potentially to account for is that the built-in display engines of portable electronic devices are notoriously poor; even if dedicated microprocessors can manage incoming bitstreams, it is a whole other architecture for spitting back out content. The same goes for the signal pathway from flash memory through a consumer BD player/PS3 and into an HDTV's HDMI input.

The signal loss caused by laying footage into an NLE and rendering out at a high bitrate is so nominal that I'm sure it's preferable to the uncertainties introduced by judging footage via playback engines that are no match for Core i7 CPUs or, for that matter, a true mastered Blu-Ray disc.

Can't say I agree with this. I get 100% perfect playback with the much more difficult to play 1920X1080 60p output directly from a Panasonic TMV700. Likewise, taking that same memory chip from the TMV700, both the PS3 and the Panasonic BD player display it perfectly. In fact, the same video on the same cruise ship at the same time with the 700, yielded artifact-free video.

The fault in terms of the NEX5 or the VG10's video is not in the output section of these devices, but rather the encoding/sensor output itself. I have zero doubt about that. It's no coincidence that nothing plays it back well. I have never seen a device that couldn't play back perfectly what it recorded...let alone external devices like the PS3 or a BD player designed for 3D, BD and card memory input.
Ken Ross is offline  
Old 09-15-2010, 04:34 PM - Thread Starter
Senior Member
 
hpmoon's Avatar
 
Join Date: Dec 2006
Location: Washington, D.C.
Posts: 254
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 10
Quote:
Originally Posted by Ken Ross View Post

The fault in terms of the NEX5 or the VG10's video is not in the output section of these devices, but rather the encoding/sensor output itself. I have zero doubt about that. It's no coincidence that nothing plays it back well. I have never seen a device that couldn't play back perfectly what it recorded...let alone external devices like the PS3 or a BD player designed for 3D, BD and card memory input.

Your zero doubt may not include a perfect render from a NLE timeline, eliminating the factor of device output shortcomings. All your previous posts continue to suggest you are leaving out that critical bench reference.

Anyway, I'd stand down from the persistent dogma that you're seeing things independently of how it compares with other devices. The only authentic testing scenario involves A versus B under identical circumstances. None of us can verify that you haven't introduced something else into the signal path, from physical presence all the way to LCD display panel, that is the source of your complaints.

I won't be blindly defending the NEX-VG10 if it sux; but I'm also very suspicious of just about everyone's technical workflow since there are so many variables in play.
hpmoon is offline  
Old 09-15-2010, 07:31 PM
Senior Member
 
Jay968's Avatar
 
Join Date: Jun 2010
Posts: 294
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 12
Quote:
Originally Posted by hpmoon View Post

Your zero doubt may not include a perfect render from a NLE timeline, eliminating the factor of device output shortcomings. All your previous posts continue to suggest you are leaving out that critical bench reference.

Anyway, I'd stand down from the persistent dogma that you're seeing things independently of how it compares with other devices. The only authentic testing scenario involves A versus B under identical circumstances. None of us can verify that you haven't introduced something else into the signal path, from physical presence all the way to LCD display panel, that is the source of your complaints.

I won't be blindly defending the NEX-VG10 if it sux; but I'm also very suspicious of just about everyone's technical workflow since there are so many variables in play.

I think I am the only one in this thread who has actually shot with the VG10 though.

As I have stated, I have gone directly from camcorder via HDMI cable into HDTV (46" Sony XBR9). I see absolutely the same results as when I create a blu-ray disk in FCP. So unless viewing directly from the camcorder via HDMI introduces something, I would have to assume that viewing this way is a pretty good way to judge its quality leaving out all the other variables.

For what it's worth, today I spent some time shooting with the 5DMKII. While that one in my opinion still isn't the greatest, the output does indeed look noticeably better than that of the VG10.
Jay968 is offline  
Old 09-15-2010, 08:18 PM - Thread Starter
Senior Member
 
hpmoon's Avatar
 
Join Date: Dec 2006
Location: Washington, D.C.
Posts: 254
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 10
Quote:
Originally Posted by Jay968 View Post

I think I am the only one in this thread who has actually shot with the VG10 though.

As I have stated, I have gone directly from camcorder via HDMI cable into HDTV (46" Sony XBR9). I see absolutely the same results as when I create a blu-ray disk in FCP. So unless viewing directly from the camcorder via HDMI introduces something, I would have to assume that viewing this way is a pretty good way to judge its quality leaving out all the other variables.

For what it's worth, today I spent some time shooting with the 5DMKII. While that one in my opinion still isn't the greatest, the output does indeed look noticeably better than that of the VG10.

One more time: there is a difference with unconfirmable variables involving both the device shortcomings of direct HDMI output, as well as the way that you import, edit and output/render your timeline. Without specifics, it could be anything. That's the point.

What's more, the proof will be in visual evidence, not subjective opinions. You can post very high-bitrate, 1080p footage at Vimeo.com, with even the raw file for users to download. I encourage you to do so at my "sub"-site, http://www.nexvg10.info.

Note that this sensor is (on paper) superior to the 5D.
hpmoon is offline  
 
Thread Tools


Forum Jump: 

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