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

[Douglas Spotted Eagle]

Try setting the vertical blur in Vegas to .002. you'll probably be pleased with that. Not that it matters, but folks working with high rez stills have been working this way for years, and it helps when delivering XDCAM vid for display on an SD monitor as well, for all the same reasons.

[Tony Tremble]

Originally Posted by Piotr Wozniacki

Again, thanks for your cooperation. We have established that the V1E/P doesn not have a flaw in 25PsF mode; I think it's safe to say its vertical resolution is just too big to be properly handled by majority of HDTVs/monitors. Well, I gues 1080/50p is the answer...

That statement is utterly preposterous. Do you really think we go to the bizarre lengths you and Steve are doing when finishing HDCAM footage from an F900?

No I don't think so. :)

I still have not heard back from my DVD supremo but I can assure you that HD television programmes are being shot in 25P on F900s and being delivered as 50i/25PsF (aka 1080i25) for broadcast. The public are not seeing twitter from HD sat broadcasts and neither will they when watching 50i (25PsF) Blu-ray media.

Do you really think the whole of the PAL region is going to have to blur their F900 footage then encode at 24 fps? I don't think so.

You are worrying needlessly and other will end up worrying needlessly.

TT

[Douglas Spotted Eagle]

Tony, this thread length has already demonstrated the lengths to which people feel the need to go for one reason or another. You've got to admit that for the most part, it's measurebating at it's finest.
There are some advantages to blurring vertically when dealing with images from the lower cost camcorders, comparing them to a CineAlta isn't really practicle nor reasonable. Yet you are correct, it's not necessary to undergo all of this for all formats, if you're going down this road at all.
But it's all in great fun, and some folks have little better to do.

This thread has remained remarkably civil given the volatility of the subject, so let's please keep it as such.

[Tony Tremble]

Originally Posted by Douglas Spotted Eagle

This thread has remained remarkably civil given the volatility of the subject, so let's please keep it as such.

I wasn't trying to light the blu-touch paper (see what I did there)! Far from it.

I think it is important that every now and then we touch base with reality in a discussion!

:)

TT

[Piotr Wozniacki]

Originally Posted by Douglas Spotted Eagle

Try setting the vertical blur in Vegas to .002. you'll probably be pleased with that. Not that it matters, but folks working with high rez stills have been working this way for years, and it helps when delivering XDCAM vid for display on an SD monitor as well, for all the same reasons.

I have tried it, Spot - and I find it to unnecessarily soften the image too much (obviously, more so than 0.001). This is why I find the finer blur control in Premiere useful - I think Sony might improve it in the future updates.

Regarding your comment on Tony's post - I could not put it better myself. The progressive picture from the V1E is way sharper than 25F from Canon, which ironically brings about the problem of aliasing or line twittering. What we're discussing here is not measurebating for its own sake, but trying to work out a method to make it watchable without softening down to a level equal or lower than the Canon's default sharpness. This is about the V1E's 25PsF only, and extrapolating it to other formats is pointless; I never meant it at all.

[Bob Grant]

Originally Posted by Tony Tremble

I wasn't trying to light the blu-touch paper (see what I did there)! Far from it.

I think it is important that every now and then we touch base with reality in a discussion!

:)

TT

Reality is that most HD down here is being shot in 50i, not 25p.
I think you'll find that by design 50i delivers less V res than 25p in most cameras due to line averaging.
Even then the broadcasters are not out of the woods. The SD downconverts can end up with twitter. Those downconverts are being done through boxes costing serious money and despite their best efforts you can still get fine detail sneaking through that causes grief.
You don't even need the V1E to create this problem, don't even need a camera. There was some great test patterns around, don't seem to be able to find them anymore so I'll have to create my own. They were of every alternate pixel black and white. Try displaying that on a 50i system, works fine a real 25p display, blows up on a 50i display. Therein lies the root of the problem.
Could Sony have taken steps to alleviate the problem shooting 25PsF. Maybe but I suspect at the expense of V res. I'd rather them keep the res and we deal with it in post, all that seems to be lacking are the ideal tools for controlling aliasing in post.
And let's not forget that we'll be adding stills, text and graphics onto our 25p timelines. Without handling that very carefully again problems with twitter will arise.

[Steve Mullen]

1) Does the V1's HDMI port have HDCP? Nope.

2) The MPEG-2 spec. allows the encoding of "field based" video either as two separately encoded FIELDS or one encoded FRAME. What does the V1 use?

The HDV spec. uses FRAME encoding for both HD1 and HD2.

Therefore, the claim that the V1 uses FIELD encoding is false and the concern that V1 progressive is compromised by not using FRAME is unnecessary. There is no difference in HDV progressive encoding between Canon, JVC, and Sony.

3) Those concerned the V1E has MPEG-2 blocking artifacts, read about what causes them in FEB issue of Broadcast Engineering. Some decoders have MPEG-2 blocking filters.

[Steve Mullen]

Originally Posted by Bob Grant

And let's not forget that we'll be adding stills, text and graphics onto our 25p timelines. Without handling that very carefully again problems with twitter will arise.

Exactly -- that's what the Flicker Filter was designed for.

[Massimiliano Minoia]

Hi everybody! This is my first post here, sorry for my bad english, I'm italian. I own the V1E since 15 Dec 2006, upgraded (?) in the early of 2007.
After the firmware fixing I noticed a V res boost switching from i to p, Why? Sony, WHY?! Someone told us that the p mode was softer than i ( the infamous oil paint effect, that is not a resolution problem but a coring/NR issue ) and Sony made the mistake of incrementing the V res causing flickering, aliasing, etc... ( the oil paint is lower but still present in fine similar color details like grass or brown trees without foliage ). I always use the FCP flicker filter workaround but I really hope that Sony will fix the problem, nobody needs a soft H res image with oversharpened V rez, twittery, flickery,more noisy, aliased... I emailed the problem to silver support, you already know the answer: try to lowering the sharpness to 3 (&*!%). Sorry again for my bad english.

[Steve Mullen]

Originally Posted by Massimiliano Minoia

After the firmware fixing I noticed a V res boost switching from i to p, Why? Sony, WHY?! Someone told us that the p mode was softer than i and Sony made the mistake of incrementing the V res causing flickering, aliasing, etc...

It's possible Tokyo created two versions of V1E firmware -- each with an error: the first firmware had the "oil paint" effect and the second with too much V. rez. To believe this you have to assume Sony engineers made two bad designs without seeing the problems that would result.

[edit]

The BBC reports that Sony downconverts HD to get SD in the A1 and Z1. The relation between the two is fixed at 2.25 -- which matches 1080 divided by 480. So this factor was designed into the V1U and V1E.

[edit]

There is a fundamental difference between PAL and NTSC cameras. PAL units must have 1.2X greater V rez. because of the difference in V rez: 576 vs 480. This is why the PAL chips always have more rows.

In order to get 576 lines, given the 2.25 factor, the V1E's HD signal needs to carry 1296-lines of rez. Of course a compromise, for example, 540 and 1215 could be chosen.

To get 1215, the only option is to raise the low-pass filter frequency to enable the capture of a signal with more vertical resolution. But, R50 HD cameras don't have extra CCD rows. The rows are fixed at 1080. When you capture a wider bandwidth signal -- with the same number of rows -- the result is inherently far more aliasing and twitter on horizontal edges/lines.

Interlace video is always sent through a low-pass filter. This is used on both U and E units. It removes line-flicker and aliasing.

This filter is not used for progressive -- so the aliasing is visible. Had Sony intended a 25p version, they would have designed the V1 differently from Day 1. Or, they blundered from Day 1. Take your pick.

To reduce the aliasing, Sony cranked up DNR which cleaned up the noise. [edit]
Of course, it also wiped-out detail. When it's video was rejected, in a panic, Sony dialed down DNR which brought back the crap. Solution, dial Sharpness down to 3. This works fine for reviews because the video is only played back.

A more well thought out solution would be to position 25p as an aquisition format that requires treatment in post after using Sharpness at 6.

[Tom Roper]

As stimulating as this thread is, there have been no measurements of anything.

[Bob Grant]

Originally Posted by Steve Mullen

1) Does the V1's HDMI port have HDCP? Nope.

2) The MPEG-2 spec. allows the encoding of "field based" video either as two separately encoded FIELDS or one encoded FRAME. What does the V1 use?

The HDV spec. uses FRAME encoding for both HD1 and HD2.

Therefore, the claim that the V1 uses FIELD encoding is false and the concern that V1 progressive is compromised by not using FRAME is unnecessary. There is no difference in HDV progressive encoding between Canon, JVC, and Sony.

3) Those concerned the V1E has MPEG-2 blocking artifacts, read about what causes them in FEB issue of Broadcast Engineering. Some decoders have MPEG-2 blocking filters.

I realise this is quoted without comment, so I'll make one.

2) Just doesn't seem to add up, there must be some difference between what Sony and Canon are doing. Why else can we not play Canon's 25F on Sony HDV kit?
Perhaps what they meant was there's no practical difference?

[Tony Tremble]

Originally Posted by Bob Grant

I realise this is quoted without comment, so I'll make one.

2) Just doesn't seem to add up, there must be some difference between what Sony and Canon are doing. Why else can we not play Canon's 25F on Sony HDV kit?
Perhaps what they meant was there's no practical difference?

Indeed Bob.

There are some _very_ significant visual differences let alone structural differences. There is a huge quality difference between the Canon and Sony encoder implementations so I would call that a practical difference.

TT

[Massimiliano Minoia]

Steve, believe me, my V1E had not V rez difference btw p and i before the firmware update, I'm sure that Sony could fix it easily, but it's a huge company, who cares about a few PAL p25 owners that wants a perfectly calibrated image?
Canon XH A1 has the V detail control, maybe the V1 has any embedded servicing menu, who knows? About SD internal downconverting I can see aliasing and twittering in 25p downconverted image, not in 50i,there is also a coarse EE in both , downconverting in FCP is a way better, but with the flicker filter engaged.
Best workflow for my little experience: HD recording with S=5 or 6, HDV or AIC importing in FCP, then applying Flicker Filter to max setting, rendering in a DVC50 timeline and exporting the movie, the resulting image has a way more rez than the Sony internal downconvert and no EE, then I check if the flickering is acceptable on my CRT, if not I apply a moderate flicker filter in the DOWNCONVERTED movie; the FCP flicker filter is a lot more effective in SD than HD because the different resolution.
Little off topic: Sony offers me a full refund if I'm unhappy with the 25p,
Canon A1 is the only alternative, but I love (and hate) my V1, if only Sony could be a better calibration...

[Mikko Lopponen]

Originally Posted by Steve Mullen

1)

The HDV spec. uses FRAME encoding for both HD1 and HD2.

Therefore, the claim that the V1 uses FIELD encoding is false and the concern that V1 progressive is compromised by not using FRAME is unnecessary. There is no difference in HDV progressive encoding between Canon, JVC, and Sony.

Not possible. Encoding a frame with fields will end up with HORRIBLE compression artifacts as the encoder will use all of its bitrates for the jagged edges. I've done some encoding tests and I seriously doubt that answer.