V1 Progressive Sharpness?

[Tony Tremble]

Originally Posted by Zsolt Gordos

Does this mean both images were taken with 25p? One looks as if it was interlaced (details are OK) and the other is smudged soft. I dont really get this if both are 25p

Both are 25P all that has changed is that a cloud obscured the sun in the second image. This is the problem Zsolt. Interlaced footage does not suffer like this. Like I said before interlaced footage looks like it has come from a different camera compared to 25P.

This shows the problem of identifying the issue. As soon as the conditions are slightly off optimum the image turns to junk. There are other issues with 25P that are really only become obvious when the image is moving e.g the excessive line twitter, clumping block noise and aliasing.

Cheers

TT

[Zsolt Gordos]

I had a longer look at the photos and realized that in the one with the details the more distant roofs also show the soft smeared effect. Basically one roof in the first row has the details, the others dont.

I would find difficult to believe the reason is the change in light conditions (however there might be some sort of circuitry or filter that is affected by light...).

Your images made me think how could this happen. It would be very interesting to see all the 25 frames taken in one second. One by one. I bet there would be difference between them in terms of softness and artifacts.

And if one could capture the 25 frames of the next second (or any other) and this pattern would show a repetitive nature then some Sony engineer might think about the reason.
All this is speculation though.

[Bob Grant]

Makes sense to me. The blacks / low light parts of the frame are where noise is the most noticable. If I was designing some form of DNR that's where I'd want it doing it's thing.
Why would you want to kill off noise, just a guess on my part but it's something that mpeg-2 encoders have a real problem with.
Now here's a thought, how is this affected by the amount of edge enhancement? Edge enhancement makes noise much worse, maybe, just maybe the problem can be tamed by lower it from the default when shooting 25p. When I say Edge enhancement read Sharpness or Detail.

What I still don't get is why this is only happening in 25p and why it's not happening in 24p or 30p.

Has anyone tried shooting these kinds of scenes on a V1U in 24p or 30p and looked for the same kind of problems?

[Tony Tremble]

Originally Posted by Bob Grant

Makes sense to me. The blacks / low light parts of the frame are where noise is the most noticable. If I was designing some form of DNR that's where I'd want it doing it's thing.
Why would you want to kill off noise, just a guess on my part but it's something that mpeg-2 encoders have a real problem with.

Are you saying that what is shown in the two images is acceptable?

The 25P sharpness level is way above that of interlaced. I shot with the default of 7 and progressive looks more like it is set to 12-15 while in interlaced it looks normal. If you turn down the sharpness the progressive footage just turns to mush below 5.

I just can't see how something like this passed any level of quality control. I have a hunch, based on the comments of Steve Mullen, that 24P and 30P are similarly affected to some degree. I reckon this might be a wider problem than just 25P.

People just don't look at an image properly or know what to expect. My dealer said he has sold loads of these cameras but there are only a few people reporting the 25P issue.

TT

[Tony Tremble]

Originally Posted by Zsolt Gordos

I had a longer look at the photos and realized that in the one with the details the more distant roofs also show the soft smeared effect. Basically one roof in the first row has the details, the others dont.

I would find difficult to believe the reason is the change in light conditions (however there might be some sort of circuitry or filter that is affected by light...).

Your images made me think how could this happen. It would be very interesting to see all the 25 frames taken in one second. One by one. I bet there would be difference between them in terms of softness and artifacts.

And if one could capture the 25 frames of the next second (or any other) and this pattern would show a repetitive nature then some Sony engineer might think about the reason.
All this is speculation though.

Zsolt

There is no repetitive problem. As soon as the cloud's shadow moves over the roof the detail is rubbed away.

All of the smudging of the roofs off in the distance is pretty normal for 25P. As soon as the detail gets fine the "filter" gets going and rubs it out. Again, this does not happen in 50i.

TT

[Michael Phillips]

It's when you pan the camera is when I notice the problem at it's worst.
Michael

[Steve Mullen]

Originally Posted by Tony Tremble

The 25P sharpness level is way above that of interlaced. I shot with the default of 7 and progressive looks more like it is set to 12-15 while in interlaced it looks normal. If you turn down the sharpness the progressive footage just turns to mush below 5.

I just can't see how something like this passed any level of quality control. I have a hunch, based on the comments of Steve Mullen, that 24P and 30P are similarly affected to some degree. I reckon this might be a wider problem than just 25P.
TT

I can tell you having shot hours of 24p there is no such problem as you say you see. And, I've posted multiple 24p and 30p comparisons with 60i -- and to date no one has reliably been able to tell which was which, except the 24P was SOFTER which, of course, is exactly what most shooting 24p want. That tells me NORMAL may be fine for both 24p and 60i. No need to dial down.

My previous posting point was that there are good reasons why I and P "MAY" look different. It makes perfect sense for them to not look 100% identical.

Saying this is NOT support for what you and a few others see. Moreover, given the 100's if not 1000's of V1Us in use -- I really doubt that no one has seen the "bug" if it existed.

In fact, I've got to wonder given the 100's of V1Es in use, why only few people have posted anything negative. Tony, it's as reasonable to assume you have one of a few bad V1Es as it is for you to assume they are all bad, but only you have the ability to see the "bug."

Frankly, I haven't looked at your pix postings because I'm only interested in how the video looks being played back. But, I just quickly looked at it. All I see is an picture with highly crushed blacks. A little too soft, but not too bad for a posted still from video.

Bottom-line -- there are dozens of luma and chroma filters that must set perectly. Seems your unit MAY not yet be precisely set.

I think it is a serious error to now try to suggest that ALL V1's have some problem based upon your single sample.

[Bob Grant]

Originally Posted by Tony Tremble

Are you saying that what is shown in the two images is acceptable?
TT

Not at all, I'm trying to fathom HOW it is happening, why a shift in the levels changes the appearance of the problem.

It also explains what Steve is saying, lots of people aren't complaining. Well sure, from what I've seen so far you need a pretty specific set of circumstances and even then perhaps only part of the frame is affected. Looking at you first image the stream looks great, the field towards the back looks horrid and the branches against the sky shows things that in motion could be awefull too.

Looking at your second two images taken in sequence shows something pretty dramatic is going on.

As to Steve's suggestion that this a one off problem. Well possibly but I'd assume all this is happening in digital land so I doubt it's a pot set in the wrong position.

So surely the answer is to get another V1E and run them side by side or else get others to check under the same conditions. In fairness to Sony as I said before if this is a genuine problem in all units I doubt typical testbench tests would reveal this problem. It's possible all units have this problem, I'd certainly have to go out of my way down here to create those shooting conditions at this time of year to try to repro the problem. Certainly I wouldn't have seen it in any of my usual camera checks.

[Alex Leith]

Originally Posted by Steve Mullen

In fact, I've got to wonder given the 100's of V1Es in use, why only few people have posted anything negative. Tony, it's as reasonable to assume you have one of a few bad V1Es as it is for you to assume they are all bad, but only you have the ability to see the "bug."

Personally, I don't think that Tony has an odd bad apple is a reasonable assumption.

It seems that all of the V1E/P users who have expressed ANYTHING on this forum have said they have the problem. I have not heard anyone saying that 25p is definately as good a V1U 24p/30p.

Anyway, I think it's safe to assume that all V1E/Ps have a problem given that Sony said to Tony that the initial Firmware update did not fix the problem.

[Tony Tremble]

The 25P issue is not limited to my unit. Sony hasn't stopped delivering units to dealers for no reason. My dealer still has not received his units back from Sony as they are not fixed. Dealers who are still selling the units are being urged to explain that there is a problem with 25P and it will be fixed under warranty.

The reason for posting the later two pics was to show how the 25P issue can suddenly raise its ugly head in a matter of seconds and how easy the flaw can be missed.

Who has ever said someone shooting progressive actually wants a softer image?

TT

[Philip Williams]

Originally Posted by Steve Mullen

<snip>...haven't looked at your pix postings because I'm only interested in how the video looks being played back. But, I just quickly looked at it. All I see is an picture with highly crushed blacks. A little too soft, but not too bad for a posted still from video.<snip>

Ok, lets keep it real. The comparative images that Tony has with the buildings.... it looks like someone took the video into Photoshop and applied the "blur" tool to litterally wipe out selective detail. Its definitely "too bad" and, to be frank, I'd have sent my XH A1 right back if it produced anything like that. You can argue about the goodnes of the V1 NTSC models, or that maybe Tony just has a bad apple (not that I think this is the case)... but I don't think it can be objectively argued that those shots are just a little too soft. They suck coming from a camcorder at this price point.

[Douglas Spotted Eagle]

Originally Posted by Michael Phillips

It's when you pan the camera is when I notice the problem at it's worst.
Michael

This is to be expected from *all* progressive cameras.
Panning and zooming are different techniques when shooting 24, 25, and 30p. 60p will be/should be more like shooting 60i, but my 60p experience is currently limited to one camera, and it wasn't all that impressive. (neither me nor the camera) :-)

[Alex Leith]

Originally Posted by Douglas Spotted Eagle

Panning and zooming are different techniques when shooting 24, 25, and 30p.

Are you sure?

I know 24p is a nightmare because of the ugly jerkiness you get from pull-down, but I've never found 25p or 50i @ 1/50th to be much of a difference. To my eyes you just get more motion blur at 25p.

[Douglas Spotted Eagle]

Yes, I'm sure.

Take your camera and do a reasonably fast pan across something that has a lot of vertical breaks, such as a picket fence, shooting 25p.

Now do the same thing shooting 50i.

You'll see strobing.

Shoot a subject moving across a landscape with high speed in both modes, you'll also notice it.
It might be that because you're so familiar with the stutter that 50i has, that it won't bother you as much, but it's definitely there. Having shot all of the modes, 50i is the most irritating for me. We used to shoot 25p fairly regularly, (or 50i) for conversion to 24p. 25p and 24p are indistinguishable.

[Alex Leith]

Originally Posted by Douglas Spotted Eagle

Yes, I'm sure.
Take your camera and do a reasonably fast pan across something that has a lot of vertical breaks, such as a picket fence, shooting 25p.
Now do the same thing shooting 50i.
You'll see strobing.
Shoot a subject moving across a landscape with high speed in both modes, you'll also notice it.
It might be that because you're so familiar with the stutter that 50i has, that it won't bother you as much, but it's definitely there. Having shot all of the modes, 50i is the most irritating for me. We used to shoot 25p fairly regularly, (or 50i) for conversion to 24p. 25p and 24p are indistinguishable.

Actually it's possible because I'm habituated to 25p. I always shoot 25p (a lot of stuff I work on is for web delivery) and I think it's about 8 years since I've shot any 50i (back in the day working for news).

Actually now you come to mention it, I do remember taking more "care" with camera moves when I started shooting 25p...

It all seems pretty natural now.