Sony Intensive Brief from HD Expo

[Serena Steuart]

Originally Posted by David Heath

No slap - you're spot on. Which is why all pro camera sensitivity ratings have to have three criteria specified - such as f8 at 2000lux, S/n 52dB.

An ASA rating only fixes the first two, so a statement such as "this camera is rated at 800ASA" is, by itself, meaningless.

The DSLR analogy is very sound - on my Canon I can make it anything between 100 and 3200ASA. To compare it sensitivity wise to another camera, you need noise level figures at any given ISO rating.

All that said, my brief experience with the EX left me feeling that it does indeed perform very well in low light. Far better than any of the current 1/3" cameras.

The EX is "rated" at 800 ASA at -3dB gain. The Z1 is 250 ASA at 0dB. That gives you a baseline for comparison.

[David Heath]

Originally Posted by Serena Steuart

The EX is "rated" at 800 ASA at -3dB gain. The Z1 is 250 ASA at 0dB. That gives you a baseline for comparison.

Not without S/N figures. They'd only give you a baseline if the EX at -3dB and Z1 at 0dB had the same S/N figures at those gain settings. That may be the case, but if not, those ASA comparisons are meaningless.

If in your example the Z1 had a S/N of (say) 56dB at -0dB, and the EX had 50dB at -3, you'd have to set the Z1 to +6dB to make relevant comparisons.

[Brian Cassar]

If in your example the Z1 had a S/N of (say) 56dB at -0dB, and the EX had 50dB at -3, you'd have to set the Z1 to +6dB to make relevant comparisons.[/QUOTE]

Interesting comment. Though I really would like to know how you have worked this out. So how can I compare my existing Sony camera (DSR 300) rated at sensitivity of f11 @ 2000lux with a S/N of 62dB with the EX1 rating of f10 @ 2000lux with a S/N of 52dB? Until I've read your comment above I thought that I will have a third of a stop less - but I'm not so sure now. Any feedback would be most welcomed.

[Jody Eldred]

An ASA rating only fixes the first two, so a statement such as "this camera is rated at 800ASA" is, by itself, meaningless.

It's not meaningless at all. If you shoot 35mm film, the ASA rating tells you how fast the film is, and therefore what light levels are required for the exposure you desire. Jeff Cree and I lit a chip chart with a specific, measurable volume of light, adjusted the iris on the XDCAM EX for correct exposure utilizing a waveform monitor, used no gain, no N.D., frame rate was 23.98P, then measured the light with a Spectra light meter and determined the camera was rated at about 800 ASA.

So for me as a D.P., I can light for the XDCAM EX the same as I would 35mm film rated at 800 ASA. If I want to use my light meter for lighting with the XDCAM EX, I can set it for ASA 800 and I'm good to go.

For those of use who rarely if ever use a light meter, the value of knowing it's rated at 800 is at the least a guide that tells us, "This camera does not need much light to make properly exposed images." 800 is darn fast, particularly when you know the F900 is below 400 (closer to 320). Shooting at 24P or 60i changes things, but the baseline is indeed useful and meaningful.

If I can find my notes and comparisons with the F900R and XDCAM HD F350, I'll share them.

[David Heath]

Originally Posted by Brian Cassar

Interesting comment. Though I really would like to know how you have worked this out. So how can I compare my existing Sony camera (DSR 300) rated at sensitivity of f11 @ 2000lux with a S/N of 62dB with the EX1 rating of f10 @ 2000lux with a S/N of 52dB? Until I've read your comment above I thought that I will have a third of a stop less - but I'm not so sure now. Any feedback would be most welcomed.

You have to put gain in the less noisy camera until the two have comparable S/N figures. Hence, given your figures, you'd need to put 10dB of gain in the EX to equalise the S/N figures, so it would then need to be reset to f16-22 to regain correct exposure.

But are you sure about the 62dB figure? It sounds somewhat high.

In practice, things get much more complicated, since prosumer cameras like the PD150 add processing as the gain is increased to reduce the noise - but at the expense of degrading the picture in other ways.

My "quick and dirty" test (in the absence of S/N meters etc) is to take both cameras somewhere pretty dark, open up the irises, and add gain to get correct exposure. Then compare pictures.

[David Heath]

Originally Posted by Jody Eldred

It's not meaningless at all. If you shoot 35mm film, the ASA rating tells you how fast the film is, and therefore what light levels are required for the exposure you desire.

It may have meaning in that sense, but not for comparison with other cameras, which is what is being discussed here. All cameras don't have the same S/N ratios at their 0dB gain setting, and sensitivity comparisons must have that taken into account to have meaning.

*IF* the 900 had an inherently much less noisy picture at 0dB setting, such that 9dB (say) could be put in without it being any noisier than the EX at 0dB, that would make it inherently MORE sensitive than the EX, even though it has a lower ASA figure at 0dB.

Exactly as with a DSLR, which can have any value from 100-3200 dialled in. How do you define it's sensitivity then? You have to specify the third parameter with electronic cameras to get like for like comparisons.

[Steven Thomas]

True, though if the EX1 is squeezing off 800 ASA at -3dB, I'd bet that it is fairly clean too.
David you are correct for wanting exact signal to noise reference to compare cameras.

I'm willing to believe Jody regarding the EX1 sensitivity.

By the way, welcome to the board Jody. You're well respected in my book.

[Serena Steuart]

There's not much point in debating in theory how one camera might perform relative to another, we just have to set them up and test. However the 800 rating has been obtained by various DPs whom I believe to be critical reviewers, so until I get hold of one and prove them wrong I don't see much to argue about. Anyone wanting really great low light capabilities will test before buying. Also there is a lot more to a good image than low noise (speaking as someone who remembers Tri-X).

[Bob Grant]

Thing is though if the camera is at -3dB what the S/N ratio is, is kind of irrelevant in terms of the ASA rating, the same as it is with film. There's nothing you can do about it. Increasing the amount of light means you've either go to iris down or up the shutter speed otherwise you're over exposing.
Now hopefully Sony haven't sacrificed S/N to get a higher sensitivity, that'd really bug me as there's nothing you can do about it. Noisy images in low light from upping the gain are one thing, noisy images in daylight are another matter.

[Serena Steuart]

Originally Posted by Bob Grant

Noisy images in low light from upping the gain are one thing, noisy images in daylight are another matter.

Agreed!! But if that happens with the EX it won't sell many.

[Brian Cassar]

Originally Posted by David Heath

You have to put gain in the less noisy camera until the two have comparable S/N figures. Hence, given your figures, you'd need to put 10dB of gain in the EX to equalise the S/N figures, so it would then need to be reset to f16-22 to regain correct exposure.

But are you sure about the 62dB figure? It sounds somewhat high.

In practice, things get much more complicated, since prosumer cameras like the PD150 add processing as the gain is increased to reduce the noise - but at the expense of degrading the picture in other ways.

My "quick and dirty" test (in the absence of S/N meters etc) is to take both cameras somewhere pretty dark, open up the irises, and add gain to get correct exposure. Then compare pictures.

Yes David - actually I've checked again and my model is the DSR-300P and it is rated as 60db, The DSR300 is rated as 62db (which I believe is the NTSC version of this camera). It is one hell of a beast when it comes to film in low light and in the past 9 years I very rarely used any lights at all - and when I did, I only used a 20 watt lamp on board powered by the camera itself (for event/wedding filming). Obviously I use gain between +6 and +9 (never more) and one can detect some grain - however the ability to film quickly and unobtrusively without lights and producing a good picture far outweighs the fact that some grain can be detected. Having said that recently I viewed some of my work (SD) on a 40' Panasonic plasma (full HD) via a PS3. The picture was awesome and the upscaling managed to hide most of the grain!

[Alan Waters]

Will leave the real techie stuff to you guys but with a new camera, much like a new film......theres not a lot of actual test footage floating around the net.

Thinking of the movies that don't get coverage before they are out.

Anyone who has seen test footage actually disapointed with the EX1?

[David Heath]

Originally Posted by Serena Steuart

However the 800 rating has been obtained by various DPs whom I believe to be critical reviewers, so until I get hold of one and prove them wrong I don't see much to argue about.

Serena - I'm not doubting that these reviewers have accurately measured an 800 rating for this camera at it's 0dB setting. But that doesn't uniquely define it's sensitivity relative to other cameras. You need the relative S/N figures for that, for all the cameras.

Not giving S/N figures is analogous to giving someone two different maps without scales. Each is accurate and useful within itself, but useless for comparing RELATIVE distances from one map to the other.

Originally Posted by Bob Grant

Now hopefully Sony haven't sacrificed S/N to get a higher sensitivity, ..........

Which is the importance of quoting S/N figures when giving sensitivity ratings! It ensures you're comparing like with like, otherwise there is no baseline.

All this said, my own quick looks at the EX make me feel that it's sensitivity performance is very good indeed, and it substantially outperforms other cameras in this price range.

[Serena Steuart]

David, agree with your technical argument and just suggesting that until we get the camera it's all a bit academic. The 800 figure predicts that we'll find it much better in low light than available HDV cameras, but until we see what Sony has built in for noise suppression and what effects that has on image quality we really are in the dark.

[Steven Thomas]

Well, in a couple weeks there will be a lot more info.

My camera should be in my hands the week of Thanksgiving. At worse, the week after.