PDX10 test
 

I'm testing this PDX10 and giving it a real workout but a couple of things have come to light. First is that the 'Display' constantly lies and is very unreliable. Don't believe me? Then try this simple test.

Zoom to full telephoto. Up the shutter speed so that you *know* the camera is in the gain-up mode (i.e. has reached maximum aperture). Film for 3 seconds. Rewind tape, play tape.

What does the readout tell you? Why, that I shot at 1/600th sec and +18dB and f1.6 - yet at full telephoto the maximum aperture is only f2.8 - a good one and a half stops slower. As a consequence I'm very suspicious of that readout. As you lower the shutter speed the display will happily tell you that you shot at f2 and f2.4 - both impossible.

Next point to clear up (lots of questions about this on the board) is that the PDX10 has three ND filters, all of them automatic in operation, all of them uncoated gelatin filters that bob in and out of the light path depoending on ambient lighting conditions.

There has been much correspondence here about aperture readouts not going beyond f4.8 even in bright sun and with auto shutter locked off, and the reason is simple. Sony simply dial in more and more ND, and if you wanted to shoot at f8 for greater depth of field at full telephoto, then hard luck.

Compared to the manual photographic tools that the 900 and the VX2k are, the 950 and the PDX10 look to be very toy-town constraining and somewhat patronising I'm sorry to say. Great pictures though, don't get me wrong, and keeping Jo Public away from diffraction losses ain't a bad thing by any means.

I have lots more to say on this fascinating camera. Not all of it good, but not all of it bad either.

tom.

Keep it coming Tom, and a big, thank you!

tom, would you please tell me how you configure out that there is ND filters inside? It seems no where it is specified , although I read from a Chinese site that TRV950 does have ND built in, but at that time I thought it might be a mistake, However waht you said reminds me of that artical I read long time ago...

If it has the automatic ND(non controlable) , then it might explain well why the readout is so misleading...I also concerned that issue and talked in that thread, though that thread seems not interesting to many people ...

Well In fact I just ordered PDX10 from J&R and will get it in two days, Tom I am looking forward to your tests, and I would like to do the same on my machine.

> though that thread seems not interesting to many people ...

Ohh not at all. I am very interested. Personally, it does not bother me that the camera has ND filters and uses them to give me a better image. I would still have liked to know that from tha advertized specs or at least the manual though... I wonder if there is some nice chap from Sony reading this... might be able to step in and clear this up.

There's nothing to 'clear up' Ignacio though like you I would have liked to know from the specification that exposure is controlled (in both manual and automatic) not by varying the aperture but by applying more or less ND. This camera is going to be used in the shutter priority mode for most of it's life and most people are going to assume that turning that exposure wheel will be setting an aperture by varying diaphragm blades.

Any photographer would be right to assume this as using different apertures can affect not only the DOF but also the way highlights and out of focus areas look. Small apertures (especially with such tiny chips) also allow you to control the amount of diffraction, so for a softer look you'd want to film at f8 and 11. I repeat - this is not marketed as a consumer cam. The price and the DVCAM logo point it at conscientious photographers.

What actually happens is that if you're filming at full telephoto in the manual exposure mode the camera has all of one stop to play with. What this means is that (contrary to the 'Display' information that's effectively useless) the camera will open to full aperture of f2.8. If it gets brighter - or you want to film at a smaller aperture - you can film at f4. If it gets brighter still the camera has no more apertures up its sleeve - it now resorts to using the first of the three ND filters. If it gets brighter still the second and then the third ND come into the optical path.

Cooleye - you're buying a camcorder with a startlingly good lens. This PDX10 has a cracker of an image and I can say this because I'm viewing it critically on a 34" Sony. Sony really have no need (apart from marketing) to pay Zeiss for their name, for this 12x zoom is bitingly sharp at very wide apertures. Thing is it has to be, you never get to see its performance at smaller apertures as the NDs won't allow it.

Ah, but you can. I've figured out a way to fool it into letting me see smaller apertures at work. More later.

tom.

> Cooleye - you're buying a camcorder with a startlingly good lens.
> This PDX10 has a cracker of an image and I can say this because
> I'm viewing it critically on a 34" Sony. Sony really have no need

Oh Tom since you are there, staring still at the monitor <grin> have you noticed how good the image continues to look if you go below 1/60 (or below 1/50 in PAL land)? I am amazed because I find I can barely see any line doubling, which is so bad with other Sony cams when you go slow shutter. Am I just using a lousy monitor or is this cam really much better at slow shutter speeds than previous models?

> Ah, but you can. I've figured out a way to fool it into letting me see
> smaller apertures at work. More later.

Jejejejej. We will all be eagerly awaiting.

Interesting point you make about the slower-than 1/50th shutter speeds Ignacio. I've done some "white diagonal" tests and like you have been impressed with the freedom from jaggies. I'm sure there's single field line doubling at work, but the DSP of the larger chips infomation feed is smoothing the results nicely.

This is indeed good news because the very poor low light performance of the PDX10 needs all the help it can get. The gain-up mode is very quiet as far as +12dB and this in combination with the 1/25th shutter speed (if you avoid the obvious pictorial give-aways) help claw back three much needed stops.

tom.

Am I then right in assuming that I shall not use any external ND filters on my PDX10?

Hans

I'll have to take you guys' word on the internal ND filters I guess. This has been discussed before and was the source of some controversy. Aside from the data code anomalies, how do you know the filters exist?

> Am I then right in assuming that I shall not use any external
> ND filters on my PDX10?

If you shoot in manual mode you will definitely need external ND filters for 1/60 second exposures. I use both ND 2 and ND 6 filters regularly in bright conditions. Now admittedly, I use them out of a desire to keep the aperture in the "sweet spot" around f4, which is the middle of the exposure bargraph readout. Tom: is it your theory that turning the exposure wheel further to reduce the aperture merely inserts internal ND filters instead of closing the iris? If that's the case then perhaps my use of the external ND filters is not needed. I would love to see some hard data on this in the form of comparison shots however.

Flickering from the ND filters?
 

Searching on the net some time ago I found someone who had dissasembled a TRV950 and had a look at the insides of it. He mentioned a two blade iris and other interesting stuff about mechanical image stabilization but said nothing about visible ND filters.

However, sometimes when shooting I have noticed some flickering, very slight, almost unnoticable. At first I though I was imagining it, but then I saw it again several times. Now that this ND filter thing has come up, I am thinking perhaps this slight flickering I saw might have been the camera adding or substracting ND to the image path. Anybody else noticed this?

Yes Hans and Boyd- you're right to think that you won't need another ND for the 950/X10. But if you wanted to shoot at maximum aperture on a sunny day you could add an ND2 or easier - you'd want to up the shutter speed and the quickest way to do this is to select the sports or portrait mode.

You say the inbuilt NDs are 'the source of some controversy', Boyd. No controversy now. 'Aside from the data code anomalies, how do I know the filters exist?' By simply looking down into the lens and turning that exposure dial - it's as plain and simple as that. Why nobody's done this in the 13 months the camera's been around I'll never know. I had one for 10 minutes and all was revealed. I told you guys the same day.

The good Sony designers have insisted (to the point of forbidding access to any apertures smaller than f4.8) that you shoot using the 'sweet spot', so there's no need to use more ND. In fact the aperture readout can be believed unless it says f1.6 and silly things like that for full tele work - the camera *is* working at f4 for most of its outdoor life.

So it's not 'my theory' Boyd, it's rock-solid fact. And comparison shots aren't going to help any - how could they? The bottom line is this: if you film outdoors with the PDX10 in auto or manual, all that's happening is the camera chooses the sweet spot (f4) and plays around with the three NDs to smoothly vary the exposure.

Next quibble. If this is so, how come the exposure wheel (actually an ND wheel for most of its life) puts unacceptable half-stop visible jumps into my footage? This is a crazy state of affairs, and is something that the VX2.1k/170 has made an attempt to correct.

The flickering is something I'll look out for, but as the NDs are mechanically raised immediately behind the two bladed iris my guess is that sudden shaking of the cam (on a fairground maybe?) could shake the NDs up and down somewhat.

tom.

> Why nobody's done this in the 13 months the camera's been around
> I'll never know. I had one for 10 minutes and all was revealed.
> I told you guys the same day.

Well there you go. After laughing my head off, I just fired up the camera, pointed a flashlight into the lens and tried confirming this. I could see the iris moving but cannot notice the ND filters... what should I expect to see exactly?

> The flickering is something I'll look out for, but as the NDs are
> mechanically raised immediately behind the two bladed iris
> my guess is that sudden shaking of the cam
> (on a fairground maybe?) could shake the NDs up and down

Ohh no... it is too subtle to be visible under such conditions. I have seen it when shooting something with a fixed frame, nothing moving at all.

<<<-- Originally posted by Ignacio Rodriguez : I could see the iris moving but cannot notice the ND filters... -->>>

OK, I just tried the same thing. Tom, I think you are right. I wasn't completely sure what I was looking at, but definitely saw something which moved in 3 steps. The color of the reflection from the flashlight also changed as this happened. All of this is indeed consistent with different neutral density filters.

Regarding the idea of choosing faster shutter speeds to force large apertures under brightly lit conditions: this is certainly an option but the higher shutter speeds change the look of the video, depending on the type of motion that is present. For example, some footage I shot from a slow moving car with trees in the foreground has an annoying strobe-like quality. Personally I prefer the motion blur of shooting at 1/60 second, and generally go one step further by processing the footage in post with an adaptive deinterlacing program that produces a psuedo 30p effect.

But this is a real revalation about how the PDX-10 works Tom; thanks for that. Next month when I finally get a little free time I'm going to have to re-think some of my assumptions about exposure and filtration. In the meantime, I've been shooting performance video of some of our operas for the first time with the PDX-10. I'm really happy with the results. It would be nice to have another f-stop or two for the dim scenes, but it really hasn't been much of a problem for me. According to the data code (but can I really believe that now? ;-) much of my footage is shot wide open at anywhere from +3db to +18db. For the most part this all looks fine. I find the noise to be most evident in the very dim blue areas however. This is somewhat similar to what I'm seeing in my digital stills taken with a Nikon 5700 (5MP 2/3" CCD).

I'm also finding that the audio sounds great on the PDX-10. Have been feeding channel 1 from the house sound board and using the mono camera-mounted mike on channel 2. This is a noticeable improvement over my VX-2000 audio using a beachtek box to connect the house board feed.

Ahhh. I saw it now. Yes I can see the color of reflections changing.

Perhaps I imagine it, but I also seem to get the idea of something turning counterclockwise, when I dial down the expsure... very strange. Could it be that the ND filters are on a semicircular support that spins and is centered to the left of the lens?

Anyway, great work Tom, Thank You for all the insight. I think it's time to write an extensive update to my review on the PDX10. If it's ok with you nice chaps, I would like to mention your experiences with this cam.

I suggested some time ago that the PDX could have variable ND filters. The concept of a wide range continiously variable ND filter would even give an extra dimension to the exposure game... One could set the shutter for the smooth motion (like 1/60) set the aperture for the DOF wanted (within the lens and diffraction limits) and adjust the ND (manually or auto) for the correct exposure. Even the large CCD cams would get extra exposure flexibility. Maybe one day...