How PsF video from the V1 is different than "p" or "F" video

[Michael Phillips]

So, Bob, can we narrow down our tests to using black text on white paper to test the 25P and see what results we get straight out of the camera. I have done some repeats of the original footage where I observed edge problem but could not repeat the results. I would like to see if all PAL owners repeated the same footage of the same test (black text on white paper) we may be able to find a common issue that we could work on.
Camera settings would need to be noted or preferably just set it to full auto. Also if we put the clip through a NLE and render it and provide feedback as to the results.
Michael.

[Mikko Lopponen]

Originally Posted by Piotr Wozniacki

Bob, pure logic says it should be done this way - just weaving when no temporal difference is detected between fields, but apparently it doesn't work this way, at least not with the majority of HD tvs and monitors.

Why?

Because the simple chips can't see the picture. They don't know what the picture is so they basically treat everything interlaced.

[Tony Tremble]

Originally Posted by Piotr Wozniacki

I'm still seriously considering to revert back to the A1, where such artefacts are simply not present (at least I haven't noticed them on my unit during testing). The picture is very "quiet", but IMHO it lacks the vividity that the V1 at sharpness 9 produces. Now that I have to make my decision, I'd be very grateful if you posted a clip from your A1 with colors as vivid and natural as those in my V1E clip, edges eqally sharp but not overenhanced - and still NO noise whatsoever. If you can do that, many of those still on the fence would benefit. Thanks!

I did create a custom profile approximating the look of the V1 but have since reverted back to less colour and edge enhancement. I guess I prefer the flatter less edgy look that the Canon gives. All the edges are sharp but just not over enhanced. That is my preference. But is it yours? Judging by your posts on this forum you do like the aesthetic of the Sony image. And there is certainly nothing wrong with that.

You are kidding yourself if you think you are going to get no noise from any camera. As soon as an image is compressed you'll get noise. The question is whether you find that noise objectionable or whether that noise is consistent. Personally I found the noise from the progressive mode of the V1 objectionable and not consistent either as Bob's detailed post explained.

The Canon does have slightly more noise to begin with BUT that noise is consistent even in poor light and in dark areas of an image. It certainly does not produce the block noise that the V1 produces.

I concur with your observation that the Canon is superior when dealing with motion.

I think you worry far too much what other people think than your own opinion. If you like the V1 image despite its obvious flaws in 25P then stick with it. Are you really going to 25P all that much? I had no complaints about the 50i mode of the V1. It was great.

Comparing the 25F mode of the Canon to the 25P mode of the Sony I prefer the slight V resolution loss of the 25F to the noisy edgy Sony. When the sharpness level is reduced to mitigate that noise an edginess then the Canon wins hands down and has no less V resolution to the V1 in IMO.

Other more important factors regarding image quality is the less DOF of the Canon and its bokeh which is gorgeous.

Ultimately image quality is subjective and what floats one person's boat might sink another. With the XH-A1 I have found I can achieve the image I have always wanted from a video camera i.e to not look like it was shot on a video camera. Other people have different requirements and tastes.

Cheers

TT

[Piotr Wozniacki]

Tony,

I get it you're not going to post any video that would be comparable to mine in sharpness and colour vividity, and yet be as quiet and free from aliasing as default A1 settings can provide. It's a pity, and it's symptomatic.

Because if I saw such an A1 clip, my decision would be obvious: say goodbye to the V1 and get the A1 for good. Otherwise, it's still a matter of a very difficult process of weighing pros and cons, and either accepting this ammount of dancing dots here and there as a trade-off for vividness and sharpness, or not.

Because I'd like to stress it again that all the other "flaws" described earlier are IMHO line twitter, which is only visible when a viewing device input is deinterlacing the progressive 1080i stream from the V1. Just returned from the local Sony retailer where I watched my video on the newest Bravia HDTV through both the HDMI and component inputs: the line twitter is simply unacceptable, just like what I get on my monitor via component. Now, why am I so sure it's coming from deinterlacing? Because the same clip played back from my PC with VLC shows no line twitter at all, and when I switch the "bob" deinterlace option in VLC, the line twitter is back - even throuh the DVI interface.

So, the question on how to paly back the progressive material from the V1 remains unanswered. I can see just one possibility, that I am unable to validate: when the progressive material goes onto the Blueray or HD DVD, the HDMI or component input on high-end HDTVs "know" they are getting progressive from 1080 DVD player and switch deinterlacing off.

Can anyone confirm this?

[Bob Grant]

Piotr,
if you can try what I did above, just straight from the camera, not tape, into the X series Bravia. Try switching the camera between P and I and see how the "twitter" looks. Try this in fairly low light on a horizontal edge.

Only reason I ask is I've only got a V series Bravia and they're none too flash, should have a full broadcast 1080 LCD in a few weeks for testing other things.

Still by my calculations my test takes the monitor out of the equation.

[Piotr Wozniacki]

Bob, I did it on the 51" X2000 Bravia, and yes it shows what your're describing in P. But it's no worse than what is happening on my 1920x1200 LCD monitor through component. It vanishes when playing back a captured P file with a software player with deinterlace off, and returns with deinterlaced on (bobbing).

[Tony Tremble]

Originally Posted by Piotr Wozniacki

Bob, I did it on the 51" X2000 Bravia, and yes it shows what your're describing in P. But it's no worse than what is happening on my 1920x1200 LCD monitor through component. It vanishes when playing back a captured P file with a software player with deinterlace off, and returns with deinterlaced on (bobbing).

Surely it is completely obvious what is happening? No?

The Bravia is not receiving or understanding the P flag being sent or not as the case may be by the V1 when you connect to it via hdmi.

When you capture the NLE can read the P flag encoded in the m2t stream hence there is no deinterlacing error.

TT

[Piotr Wozniacki]

Originally Posted by Tony Tremble

Surely it is completely obvious what is happening? No?

The Bravia is not receiving or understanding the P flag being sent or not as the case may be by the V1 when you connect to it via hdmi.

When you capture the NLE can read the P flag encoded in the m2t stream hence there is no deinterlacing error.

TT

It has always been obvious to me, while you kept saying that the "flaw" has nothing to do with deinterlacing, Tony.

Anyway, I'm compiling a short DVD with BlueRay-formatted, 25p clips from my test V1E and will give it a try at the retailer's on his Bravia. If the twitter is gone, for me the V1E is the winner as I never intended to watch my video straight from the camera, anyway.

[Tony Tremble]

Originally Posted by Piotr Wozniacki

It has always been obvious to me, while you kept saying that the "flaw" has nothing to do with deinterlacing, Tony.

The real flaws are not those introduced by HDTV deinterlacing.

Bob Grant has posted the most complete appraisal of the V1 progressive issues and I wholeheartedly concur with his observations. Let's not forget Brett Sherman's red shed clip which shows the same flaws.

I can see the same flaws in your footage you posted. My monitor is not deinterlacing and neither is the software. I can see aliasing aplenty. I can see the oil paint effect in the dark areas of the garage and I can see the dancing noise round contrasty edges.

TT

[Piotr Wozniacki]

Tony, I am more than aware the picture is far from perfect. Could you please indicate more precisely where exactly you can see the oil paint?

However, the ultimate test will be for me that of trying how a Bravia HDTV can interpret a progressive material, put on a BlueRay player.

Therefore, I'm in an urgent need to put some HDV 1080/25p file (around 2 GB) on a regular DVD (don't have HD or BlueRay burner) in such a way that it be readable by a BlueRay player. Is it possible? Which authoring application will allow me to create the structure on my HDD, so that I could just copy it to a DVD as data? I've heard this is possible with Ulead VideoStudio, but only with HD DVD. What is the BlueRay folder/file structure, and what file format is playable (it's \HVDVD_TS\*.mpg for the HD DVD). Please help!

Regards

Piotr

[Steve Mullen]

Originally Posted by Tony Tremble

My monitor is not deinterlacing and neither is the software. I can see aliasing aplenty. I can see the oil paint effect in the dark areas of the garage and I can see the dancing noise round contrasty edges.

TT

The ONLY way your monitor is not deinterlacing is if its an interlace CRT display. Since almost no one has HD CRT monitors at home or in public -- the "issue" is no issue.

The question is what P mode looks like on a Bravia or other high quality flat-screen monitor. At 3 the line twitter and aliasing are gone. They are gone in Europe and Piotr's video played in the USA show no signs either.

The only open questions are "is 3 too soft" and if so where is the point where aliasing looks worse than a too soft pix. You can't answer these questions. Piotr can.

I can tell you the V1 offers equal H. rez to the Canon and much much better V. rez than the Canon in 24F. Even when the Canon is in interlace, the V1 matches it.

PS: Another week of V1E sales and you are still the only one posting they can see the supposed oil paint effect on fixed units. Simon no longer does, nor was it reported in the dvinfo UK review. And Piotr's video shows no signs either.

Of course if you bring up a frame in PS and blow it up 4X you can see a difference, but who does this?

[Tony Tremble]

Originally Posted by Steve Mullen

The ONLY way your monitor is not deinterlacing is if its an interlace CRT display. Since almost no one has HD CRT monitors at home or in public -- the "issue" is no issue.

ALL the artefacts are clearly visible viewing the clips viewed in QT Pro and VLC displayed on a 23" Apple Cinema Display and also on 24" Dell. Both displays are 1920x1200 showing the full 1440x1080 HDV.

Neither display are deinterlacing the content and neither VLC nor QT are deinterlacing it either.

If you cannot see the artefacts in Piotr's clip then I can only assume you have not viewed it or are incapable of viewing it correctly yourself.

Until I am told otherwise I shall counter your arguments at every opportunity as they do not reflect accurately the reason for poor progressive performance from the V1.

Have a great weekend.

TT

[Steve Mullen]

Originally Posted by Tony Tremble

ALL the artefacts are clearly visible viewing the clips viewed in QT Pro and VLC displayed on a 23" Apple Cinema Display and also on 24" Dell. Both displays are 1920x1200 showing the full 1440x1080 HDV.

Until I am told otherwise I shall counter your arguments at every opportunity as they do not reflect accurately the reason for poor progressive performance from the V1.

Have a great weekend.

TT

Likewise.

I'm not claiming line-twitter aliasing is CAUSED by the monitor! I'm only saying a monitor can create artifacts when none are present and remove artifacts that are present. You can see that by switching between the various VLC modes.

I have a very good idea why the Region 50 units need a setting of 3 to virtually eliminate line-twitter and aliasing. At 3 these artifacts are no longer a problem -- hence THEY cannot cause any "performance" problems.

Since P mode is 99% used for a film look -- the softer look at 3 matches the goal of film look.

Now IF you consider the softer video at 3 to be poor performance -- that's your option. I'm not going to fight over this point since I want a high detail pix. But, i would use 50i.

Have a good weekend.

[Bob Grant]

Originally Posted by Steve Mullen

PS: Another week of V1E sales and you are still the only one posting they can see the supposed oil paint effect on fixed units. Simon no longer does, nor was it reported in the dvinfo UK review. And Piotr's video shows no signs either.

Of course if you bring up a frame in PS and blow it up 4X you can see a difference, but who does this?

Sorry but I cannot find any post from Simon saying he no longer sees the problem. His last comment on CML (that I don't have access to) seems to be that the excessive coring issue has been reduced in the V1E post the fix.

I've had one email from a V1U user saying he's returned the camera because of the oil painting effect that he can clearly see.

On the V1P, I can see the problem also.

As for another week of V1 sales, well I have no hard evidence to back this up, no one does, but I seriously doubt this camera is selling in anything like the quantities the Z1 has. Most of the early adopters have already bought a Z1, those that wanted P probably already have a suitable camera.

And out of those that have bought a V1 and seen the problem, well they're hardly being welcomed with open arms here. So far they've been told screen grabs mean nothing, they don't know what they're doing, their monitors are at fault etc. You'd have to have a pretty think hide to stick you hand up around here.

[Bob Grant]

Just a further interesting note.
The HDMI output from the V1P is capable of rescaling, it would seem the HDMI spec permits a display device to define the resolution it needs and the sending device will rescale. So it's quite likely when you connect the camera to a HDMI display the camera's HDMI output is being rescaled by the HDMI chip in the camera to match the resolution of the display device.
That could maybe, perhaps, explain a few things.
This was something I never knew until we tested the V1P with another device that we wrongly thought had a rescaler, turns out it doesn't need one, it forces the camera to do it.