TRV950 -- various questions

[Boyd Ostroff]

<<<-- Originally posted by Tom Hardwick : While the push auto button is depressed the camera will be in auto focus mode; once released you’re back to locked focus. A focused measurement in metres appears for about three seconds in the viewfinder -->>>

This is really a nice feature on both the PDX-10 and TRV-950, and is how I shoot most of the time. Recently I've been using the autofocus push button, then noting the distance (in meters) that displays in the viewfinder. Then I can manually focus with the ring later, and use the distance display to hit the mark. Unfortunately the focus ring is not very responsive, but you can get the hang of it after some practice, and the distance readout is a real plus as it gives you a little feedback.

[Shawn Mielke]

Is the focus ring on your vx2000 significantly more, less, or similarly responsive compared with the PDX10, Boyd (or Tom)?
To add to the pot, I haven't a great deal of experience with many different cameras, but I find the PDX10'S autofocus to be very responsive, with very little hunting or foolery. I really have yet to see it become confused between two objects. Just lucky, I guess. This camera has an optional in- frame target display frame that is basically brackets, [ :-) ], sorry, can't remember what it's actually called, that I find extremely helpful to my eye in helping define the proportions of the image, and I think also they can give a sense of where the autofocus is going to land. I'm sure I shoot what is called "run n gun", for the most part, in a classroom setting, where I'm tracking a teacher/speaker, and also covering a room of student/teacher trainees, picking up moments between two people here and there amidst a lot of moments, from across the way, with, sometimes, a bit of someone along the side of the frame, as part of composition, and the autofocus is almost never confused.

[Boyd Ostroff]

Well it's just "different" than the VX-2000. It's a smaller, tighter ring. If you turn it very slowly it doesn't seem to do anything, which is reinforced by not seeing the focus distance change in the viewfinder. Neither camera has a decent focus ring unfortunately, but I'd give the PDX-10 the advantage due to the distance display, so at least you have some idea of what's happening, and some hope of hitting the target in a rack focus situation.

The main issue I have with autofocus is that it gets lost under very dark conditions. Generally, these would be conditions where there isn't enough light for a decent image anyway. But where it always gets me is when I film our performances. When the houselights go out and the theatre is dark the autofocus either hunts or becomes totally blurred. Then when the stagelights come on it snaps back into focus. I really hate this sort of look right at the beginning of a scene, so I now make it a point to be sure to have a rough focus laid in and switch to manual before the houselights go out. Then as I film I continue to use manual focus, augmented by the instantaneous auto pushbutton.

[John Jay]

of course the big booty with the focus in low light with the 950 is the advantage of the laser rangefinder (figure movie Predator) which is happenin in photo mode and places a matrix of red lines on your talents forehead if u like :)

- you can also skwirt them with 5 multipulses from the flashlight and watch them run for cover (even funnier)

when focussed - switch to cam mode and shoot

- pure terror they think the SWAT team is in da house :)



on topic

TIM - if you search through my posts you will find some footage you are seeking

[Shawn Mielke]

Beg pardon?

[Tom Hardwick]

Good points Boyd, and I know what you mean by the auto-focus hunting as the stage lights come up on scene 4. The PDX10 overcomes this in that you can preset the focus (in the dark) by looking at the 3 seconds of 'distance display', and this is much better than the VX's 'where the hell are we?' system.

Both focus rings do nothing if turned too slowly but the PDX's at least warns you that nothing's happening as no distance display appears.

But the real cruncher is that because the VX sees so much better in the dark it's autofocus does likewise. The three stop difference is a massive knockout.

tom.

[Patrick Nee]

I got my 950E about a year ago and have been delighted with the results it can deliver. I am slightly disappointed with low light performance (learnt to live with it) but what really ticks me off is the ability of the PDX10 to reproduce real 16:9 where as the 950 crops. As I understand, there is no difference in the CCD's of these cameras and technically speaking the 950 could do 16:9 a la PDX. So my question is, does anyone know of a hack that would enable true 16:9 on the 950? It sounds that the only thing that differentiates the two is the way in which the cameras are programmed to scan the CCD in 16:9 mode, so could the 950 be reprogrammed to do this or is the issue more complicated?

[Ignacio Rodriguez]

There are also other differences: the PDX10 comes with an XLR audio input module and what seems to be a hypercardoid external condenser microphone. I think the 950 has a color VF and not the black and white one which comes with the PDX10. Also I presume (though I could be mistaken) that the 950 does not have a custum setup menu nor can it record in DVCAM mode (though it most likely can play back DVCAM tapes).

However, I do think you are probably right in assuming that the 950 is physically capable of using the full CCD width for 16:9 video. The firmware is probably responsible for this.

Two questions: on your 950, do the angular width and hight of the image change (get wider and taller) when you switch from video mode to 'memory' mode? When you are in memory mode, does the camera output pseudo progressive 30fps video through the firewire port? If the answer to both questions is yes, then the video processing of the cam is very similar to the PDX10 and you are probably right in your assumption.

I assume you already checked whether the image field goes wider when setting video mode from 4:3 to 16:9. If it does, then you have nothing to worry about: you are in 'native' 16:9. If it does not, then you have a good reason to get mad at Sony for deliberately crippling the camera just so they can charge more for the 'pro' model.

[Patrick Nee]

You are right Ignacio, the PDX, obviously has DVCAM capability, XLR and the VF is B&W. It also lacks the flash that the 950 has as this area is used for the XLR mounting shoe; I personally would rather sling a beachtek under the cam as I imagine the XLR makes the cam real front heavy. The 950 does have the custom preset however, but does not do stuff like continuous TC and does not come with the bid hood that would probably work with my WA lens. (anyone know where to get a spare?)
But what really gets me is that it would seem Sony has cheated the prosumer out of true 16:9, reserving this for the PDX. The 950 has the same 1/4 inch CCD as the PDX and in 4:3 they both only scan 1/4.7. They both use the full 1/4 inch in Memory mode, hence the wider picture. But in 16:9 the 950 crops and looses resolution where as the PDX uses the full CCD as in memory mode. I personally fell cheated by this as it seems to be a conscious decision by Sony to reserve this feature for the PDX when the 950 is just as capable of doing it on paper. I just hope that some genius comes up with a way of hacking into the 950 firmware to enable true 16:9.

[Tom Hardwick]

You shouldn't feel cheated Patrick, you chose the flash gun, others chose the 16:9 option. And you paid a lot less than the PDX guy for the same imaging lens/chip assembly.

The PD100 wasn't much of a sales success alongside the much more popular (and much cheaper) TRV900, and Sony learnt from this. They've now given the PDX10 an USP that even makes potential VX2000 buyers sit up and take notice.

But like you I think it can be hacked. The very fact that the stills utilise the full chip width points to the fact that the 950's black bars are just a way of hiding what's actually going on in the background. Think on this: The 950's black bars are simply vertical (pillar box), hiding the full width and wider-angle view of the 16:9 capability. How's that for a conspiracy theory?

tom.

[Boyd Ostroff]

Patrick,

This has been discussed in the past, but I don't know that there is a definitive answer. I can't find the link, but a year or so back somebody did a few interesting tests to try and determine how many pixels were actually used in the various modes on the PDX-10 and TRV-950. Basically what he did was take a still photo (in photo mode) since that uses all the pixels. Then, without moving the camera he shot some 4:3 and some 16:9 video. By comparing these with the still photo he was able to divine how much of the CCD was being used.

The results indicated that the PDX-10 did expand the field of view and use enough pixels to acheive a "real" 16:9 image. The TRV-950 used a smaller section of the CCD, however it still appeared to use more than 480 vertical lines, and did not crop to 360 like many other camcorders.

So the consensus was that while the TRV-950 doesn't use as many pixels as the PDX-10 in 16:9 mode, it should still surpass the vertical resolution of cameras like the PD-150 or VX-2000. But I should point out that others questioned these findings as well.

Whoa... hold the phone a minute... I just dug through some old postings and find the link in question: http://www.techshop.net/PDX-10/!

[Ignacio Rodriguez]

Really? Vertical black bars in the VF? I would think Sony would at least be a little more subtle!

Intersting note Boyd. I would think other megapixel DV cameras might also offer better 16:9 performance even if the image does not visibly widen when changing modes.

[Tom Hardwick]

The link you posted shows what I've found Boyd - that the PDX10 doesn't use the full 4:3 vertical resolution when put into the 16:9 mode. Why Sony have 'degraded' the resolution potential of the 16:9 image I can't imagine.

Graph paper filming tests show that the 16:9 mode uses 10% less vertical CCD pixels than in the 4:3 mode but 20% more horizontal pixels. In a perfect world the vertical resolution would remain unaffected and the horizontal pixel count would increase by 25%, but this Sony solution is most certainly a step in the right direction. Low light sensitivity, Steady Shot and camera control are all unaffected by the switch to widescreen, though the telephoto reach is somewhat reduced and the slight barrel distortion visible at the wide-angle end of the zoom becomes more apparent.

tom.

[Tom Hardwick]

You say:

Now if there were a nice algorythm to get rid of the vertical smear, we would all be happy!

But surely, Sony know how to fix this fault. Neither the PD100 nor the PD150 smear in this way, and both of them are older designs. So what's the common demonitor? Why, bigger chips. Easy, huh? The smear seems a high price to pay for the mega-pixel capability if you ask me.

tom.

[Bob J. Trimmer]

I thank you all for this discussion, I have been interested in this myself. Another interesting thing, I also have a Sony 510 Digital 8. It seems to have true 16:9. When I turn it on the image is much wider. I also have the Canon Optura PI. In the 16:9 setting the image is taller. My 950 is as you described, just bars across top & bottom. The 510 D8 seemingly has true 16:9, but also has Bars top & bottom. The Optura PI produces a taller image in 16:9. I don't understand how it is able to convert to a wider image.