I think the focus location is slightly 'deeper' into the scene with the progressive.

i.e. The middle of the car is sharper with the progressive but less sharp on the front tire (compared to the interlaced).

Both captures look reasonably good though............. with the cautionary note of the mild CA.

Thanks for posting
JohnG

Simon

I don't suppose it would be possible to post any footage? Something with movement and detail?

Cheers

TT

I'm not using 24p or pulldown. This was the European model and I was testing 25p. The default in Vegas for this is Upper Field first. I have to manually set the clip properties for the progressive footage to "progressive".

There is no issue with Vegas and this camera. Unlike the clumsiness that is 24p, 25p does not require any special handling other than telling Vegas that it is progressive.

I can assure you that focus has not changed in any way or form. The only thing that changed was my switching between progressive scan and interlaced. Camera was in full manual operation and my fingers didn't go anywhere near the focus.

Look, there are huge chunks of detail that are missing from the progressive images that are not explainable even if Vegas had been set incorrectly (which was not the case).

??

Flipping back and forth I see no loss in detail, but rather a change in the sharpening settings.

On the roof edge for example, the interlaced image has a distinct detail-less black halo. The progressive image has a smooth transition to the background. Same as in the driveway, the sharpening enhances the edges of the stones.

It's also difficult to know if you're on an i-frame or not. Who knows how the MPEG-2 compression enters into it?

-Steve

I would say that people finally need to start sharing footage. They obviously are not allowed to... but that is what I want to see right now. All this endless analysing of images. It's all about how it looks in motion.

But Vegas 6.0d has no preset for HDV 1080 25p on my machines, only 720. And I'm still unclear if field order should be "none" for this camera (like it is for the DVX, XL2, HVX etc in progressive modes). If you go into Vegas and chose the HDV 720-25p preset, it sets the field order to "none".

This all may have no effect on the footage, but I'm pretty sure Vegas will need some updates to fully support this cam.

Screen grabs are always tricky to judge - these kind of discussions happen with every camera released. Unless Sony is lying about the cameras shooting mode, progressive mode will be as sharp and usually sharper than interlaced in all situations but some firmware or other issue may explain what you are seeing.

There's no preset for 25p, but all you have to do is select progressive scan in a 1080 HDV project properties. I am rather familiar with working with 25p footage in Vegas! Just because there is no preset, it doesn't mean you can't do it. Presets are only what the programmers thought you would need. There is no "None" setting for field order in Vegas. Only Upper Field, Lower Field, or Progressive.

Vegas does not need any updates for the V1. Not for 25p anyway. Unlike 24p (and this is something I really need to emphasise) 25p does NOT need any special handling other than telling Vegas that the footage is progressive. There are NO special requirements for 25p. None. Nada. Zilch. In fact you could edit it as interlaced footage and it would still render out correctly because each alternate field is part of the same frame. No pulldown or other such clumsey muck is required.

These are very easy to judge. There is huge loss of detail, period. If you watch the camera feed live, as I did on a 32" high def CRT, there is a loss of detail and resolution that is very apparent.

I really don't know how some of you can only see a slight softness in the progressive image. Are you guys actually looking at these grabs at their full size, or with a scaled Windows picture viewer or something?! In fact I find it utterly unbelievable that some of you can't see the total and utter loss of detail. The detail is either there, or it isn't. And I can quite clearly see the detail dissappearing on an A/B switch. I frame or not, this is a totally static frame. The detail on the driveway does not simply go 'soft'. Whole swathes of it completely dissappear.

There should be absolutely no difference at all, none, between the screen grabs if the progressive mode was working as advertised.

I'm looking at it at full resolution on a 1920x1200 monitor. Not only do I see a difference in sharpening, I see a difference in position. If you're serious about claiming a loss of resolution, ask yourself the following:

Why is there a change in position?
Was the camera locked off?
Was there camera shake?
Is the loss of detail due camera motion blur?
Could it be that the loss of detail is macroblocking due to camera motion?
Where in the compression scheme is the frame taken from?
What were the in-camera sharpening settings?
Does applying sharpening after the fact achieve the same result (answer: no. compression noise and blocking, in addition to different algorithms contribute to the problem)

The list goes on. This is a shoddy experiment from a scientific standpoint (no offense meant - thanks for posting grabs!), and if you're going to make claims on camera performance, it's better to state all parameters.

-Steve

It just the wording we are both right (see attached) field order settings in Vegas 6 and 7 is "none(progressive)" or "Upper Field or Lower Field". I'm also aware you can change the presets - I do all the time. Since you have 25p version of camera, it may well be that Vegas fully supports the camera right now.


Evidently they are not, otherwise there would not be the varoius opinions here. I've looked at the graps on mutiple hi-rez monitor at full size and I see the progressive as sharper at the point of focus and softer in the out of focus regions.

Vegas does fully support the 25p HDV format right now, very different than the 24p.
Bear in mind, Simon says he's viewing on a 30" monitor. That is not native size for 1080, therefore there is absolutely some scaling involved. Between all the various monitors, it's quite possible various people are seeing various results. PNG is the best way to look at these, of course, but if you can accept stills (which really is a poor way of looking at motion pictures) then some zooms might be appropriate.

That said...lighten up, guys. At the least, you're getting to see early on, footage from these camcorders. I don't see anyone bitching about the Canon, JVC, or Panasonic footage...As someone who spent a lot of time shooting various images for sharing, it sorta makes one wonder "why am I doing this" when people are arguing about what you did or didn't do. Maybe Simon is doing an article on the camera for a magazine as I have done, and that requires caution as to what is posted, because a magazine has purchased the rights to specific images.
Either way...a tad of appreciation to Steve Mullen, Simon Wyndham, and anyone else who has posted information might be in order, rather than the roasting and argument?

Actually, people have been bitching about the Canon footage too. That's what people do any time there's any new camera and people post frame grabs.

"Bitching" ?? Really?

I thought people were very grateful to Kaku Ito for all of his work.

Of course that gratitude also goes to Simon, Steve and DSE for making information, stills and clips available. I think it is a misreading of the level of interest that these two cameras have to say people are bitching. I thought it was good healthy discourse.

All I can see from this thread is a bit of brainstorming going on as people try and understand the reason for the progressive footage resolution loss and not trying to find fault with Simon's method.

Does Vegas know the difference between 25P and 25PsF?

Clutching at straws....

TT

Hi Doug. I have viewed the footage on both a 32" HD CRT (1080 native, no scaling), and on my video editing LCD's at the 'full' resolution in the Vegas 7 preview window.

Thats exactly how I feel at present! Along with the annoyance that people are assuming I'm making newbie style mistakes when setting up my NLE, or that I don't know how to use it properly!

Yes, I am reviewing the camera for Showreel.

Steven, I've gone through this several times. I'll go through it once more.

1. Yes, the camera static, resting on the floor.

2. No there was not any camera shake. Not unless there was an earthquake in the area that shook the ground that the camera was resting on.

3. No, the loss of detail is not down to motion blur. The camera was not moving.

4. No there is no macro blocking due to motion (when have you ever seen macroblocking that looked like that?). The camera was static. I'll say again, the camera was static.

5. How would I know? The camera was static. Its not like an in motion frame where such things can be picked out.

6. Sharpening settings were on the default (7) for BOTH shots. Actually, even with detail turned right down the camera is still sharp.

7. Why would I want to add sharpening afterwards?

The camera is in a VERY slightly different position. I moved it slightly while pressing the menu buttons. But the camera was NOT moving during the shot one iota. And a slight shift in position doesn't change the fact that a load of detail is missing.

http://www.simonwyndham.co.uk/Sony%20V1U/blowup.png

Try that. Now, if I was face to face wit you, could honestly and truthfully tell me with all your heart that the progressive version of the image is only 'slightly softer' than the interlaced one? Really? There can be no dispute. That progressive frame is seriously messed up resolution and detail wise. If there are varying opinions after seeing the shot above, then I give up all hope, and will have to wonder forever if we are all living on the same planet!

There is no change in depth of field. The camera was on its widest angle with the focus locked off to infinity. And this isn't exactly a 35mm film camera, so any idea that the depth of field can be manipulated selectively, especially to the degree of difference that there is from that angle is utterly absurd anyway. I'll also add that the interlaced footage was taken after the progressive one (so you can forget any idea that I knocked it out of focus for the proscan shot to begin with).

Vegas treats them the same. As soon as you tell it that footage is progressive, it combines the two alternate fields to make a full progressive image to display it fully in the preview window. The difference in the preview window when you tell Vegas that footage is progressive is instantly noticeable as a higher detail/resolution image. Because of that I know absolutely that vegas was handling the footage correctly.

Even if it wasn't, it still wouldn't explain the 'paint filter' effect. As you can see from my last post with detail blown up 300%, people who see no difference must need glasses!

Evidently they are not, otherwise there would not be the varoius opinions here. I've looked at the graps on mutiple hi-rez monitor at full size and I see the progressive as sharper at the point of focus and softer in the out of focus regions.[/QUOTE]

The clue is in the noise. Take a good long look at the noise in the black of the gold car's rear tyre.

In the interlaced image the noise is high frequency. It should be identical in the progressive image which it most certainly isn't! It has nothing to do with focus shift. Even if the focus had shifted, which I don't believe it has, the noise would be high frequency and the same as the interlaced all things being equal. The noise is inherent to the system and will be the same for either format and independent of focus.

So that means the noise has been "filtered" either in-camera or somehow during capture.

As we know this is a pre-production model so it is plausible that not all functions have been turned on or might even contain bugs. To my eyes progressive image looks like an interlaced clip that has been crudely deinterlaced and then smoothed.

TT

Exactly Tony! Finally someone who is looking properly!

It wasn't filtered in capture. It was just a straight Vegas 7 HDV capture. On a live display from the camera to a high def display the loss of resolution in proscan mode is noticeable there too.

Exactly.

I'd like to know how anyone can determine where the point of focus is on an image with such a huge DOF!!

:)

TT

Especially when the point of focus was infinity!

To me it looks like a little too much noise reduction going on....

I hope the camera just needs a little more tuning in the on board processing
engine before it ships....

Some one needs to send this to Sony and ask " what the hey "

Simon, I'm unawares of a CRT that displays all 1080 lines, and at 30 inches, it's scaled somehow. The Sony BVM series only display 800 "true" lines, which is why we use it for HD/SDI matched to a 2k projector, and try to discern both.
However, your screen grabs do show compelling differences that we're not seeing with the footage we had from the V1U. Makes me wonder if the 25p model is significantly different somehow?

I live in a PAL country, so I am interested in clarifying the issue (if any) apparent with the V1E version. I am watching Simon's grabs on a 1920x1200 monitor full screen and while I'm not entering the dispute on focal point, possible camera shake compression artefacts and alike, I can clearly see the loss of resolution and absence of some fine detail in the progressive picture. However, those flaws are so evident and serious I find it hard to believe they actually origin in the camera itself. With all due respect, Simon, there HAS to be some other reason. Sony don't employ idiots, who would market a new camera as truly progressive, while in fact it simply deinterlaces and smoothes interlaced material with some coarse on board magics.

Simon,

If you can find this scenario replicating itself in other screen grabs then I would think you have brought light to some serious issues with the camera. I am thankful for the people like you that are sharing their findings. Hopefully Sony's progressive scanning will not prove to be inferior. I think it sounds promising that DSE has not seen a loss in quality on his end.

It's also pretty important to note that these are *all* pre-production models. Consider them Beta units if you will. I had two different units that were quite unique in what they had completed. For instance, the first unit had a smooth handgrip finish, no interface for the DR60, no timelapse, no 24p. The next one to come in had all those features, but a couple other menu options were not complete. So...it may well be that the V1E that Simon has is in that same category.

Viewing full resolution
 

Hi DSE,

You mentioned needing a monitor over 30" to see full 1920x1080, but I know for sure that both the Apple 23" and the Dell 24" computer LCD monitors, while not showing proper gamma, will show exact 1:1 detail of an 1080i HD image. So for detail viewing, they are great. I own the Dell and it is showing full resolution.

Thanks for all of your postings. We are so much the better for it.

-Christopher

Thanks Doug. Okay you may be right about the scaling. Even still the resolution drop was still very apparent on that monitor. And as you have concurred, there is definitely something wrong with the image.

I'm not sure. It did seem pretty final. Timelapse etc was all there, and there didn't seem to be any missing features that I could tell. Although it was a pre-production model and anything is possible.

When I received the camera the weather was terrible and I couldn't go out and about to try it out, so I apologise for the following shots. They were taken very quickly literally minutes after I took delivery of it. So don't take them as an example of my testing!!!! But regardless of the crap subject matter and piss poor composition, take a look at the detail The brickwork on the garage and house for example.

http://www.simonwyndham.co.uk/Sony%2...interlaced.png
http://www.simonwyndham.co.uk/Sony%2...iorProscan.png

With all due respect, there isn't.

I do think that this is an E model problem only.

as I said with the previous two grabs: the quality loss too evident for Sony to not have been noticed, or pretend it's not there. This particular camery must be faulty, as simple as that.

Thank you Simon very much for your effort; I can imagine how disappointed you must be with the camera and - with all this discussion going on, where some people don't see the obvious - I share your frustration.

But you know what? I think the good news is that - like I said - it's too bad to be there in the final product. Let's hope so!

Thanks Piotr.

Funnily enough though I am not dissapointed with the camera. If this is an issue I am sure Sony will sort it out, so these shots are only my findings with the camera that I had.

In all other respects I liked the camera a lot. But for my full findings you'll need to read Showreel :)

With all due respect, they *MAY* be another reason. I don't think being quite that dismissive of polite posters like Piotr and DSE is awfully helpful or condusive to a calm discussion IMO.

Is there anyway with your conacts with Sony or Douglas' contacts we could get to the bottom of this?

It would be nice to close the subject with fact rather than conjecture.

Cheers

TT

I sure will! I have just got rid of the HC1 in expectation for the V1, but - having been with Sony for some 30 years now with all sort of electronic equipment - must know for sure it's at least as much better than the Canon XH-A1 as more sexy it looks:) To me, of course!

The only thing I can think is that this is a contrived solution by Sony so this potentially superb little camera doesn't take sales away from the XDCAM HD big bros.

Call me cynical....!

But that doesn't ring quite true because the 24p of the V1U doesn't suffer the same fate.

TT

Like?

The data in the picture is the data in the picture. No more, no less. This isn't an analogue capture where filters etc can be used, HDV is a bit for bit transfer over firewire. And as I have mentioned before, 25p does not have the same special requirements that 24p does. I work with 25p footage all the time (pretty much exclusively actually).

As I have mentioned before on this thread on more than one occasion, the softness in the progressive mode was very apparent when viewed live from the camera on a high def monitor, let alone after capture. This is a camera issue. Not a NLE one, or something that I may have done. There simply isn't anything that I *could* do on capture and display of this footage that would cause this result.

OK guys...I think this particular horse has been beaten to death.
There is obviously something not explicable at this time regarding how the V1E is seeing the differences between progressive and interlaced captures.
Neither Steve Mullen or myself have seen this with the V1U.
All camcorders tested are pre-production units without regard to how "complete" they may or may not be.
Simon is a professional cameraman, I'd consider myself in that same category although I'm primarily an editor, I spend a LOT of time behind the lens as well, particularly in Xtreme sports. Trying to suggest that Simon doesn't know what he's doing is non-productive; he's got credits and chops that are exceptionally respectable.

Let's table this discussion for now, and wait til we have closer-to-finished camcorders with which to have a truly informed discussion/dialog.