F3 Light Sensitivity vs. HVX200

[Matthias Sachal]

Hey Guys Ive got a question. I just tested the F3 against my HVX200 in terms of Light Sensitivity. I set both lenses to f 3.5 - shutter 1/50 and no gain. Isnt the F3 supposed to be a LOT brighter than the HVX already without any gain? Its only a bit brighter (not even a whole F Stop) and of course theres less noise but is that what everyone means when they talk about the sensitive F3 Sensor or did I miss something? (ND Filter was of course, off.)

[Jean-Philippe Archibald]

I don't know, but I like your studio!

[Brian Drysdale]

Are you using the same curves? Different curves can change the ISO rating of a camera, There's a lot more shadow detail in the F3.

[Dave Sperling]

Were both cameras set to 50i?
There are various ways in which the F3's chip may seem to exibit a greater or lower senstitivity -- enabling certain of the cine gammas, for instance, or adding the shutter.
To perform a more accurate comparison, set both cameras to 50i (or 60i here in the States), turn off any gamma or matrix settings, and turn off the shutter. Also put a cine-style prime on the F3.
Are you using one of the Sony prime lenses that come with the F3K package? If not, (or if you're using a zoom) the f/stop on the lens (which should be accurate for computing DOF, but not for computing exposure) may be significantly different than the t/stop of the light actually getting through the lens. That's why cine lenses come with a T/stop, which in the case of a zoom is often a half stop slower than the f/stop (my recollection from the old film lenses I had that displayed both f/ and t/ stops.)
Of course the cameras are really two very different beasts, with the HVX 200 essentially having a standard-def-resolution chip (hence large pixels) that's up-ressed electronically to create an 'HD recorded image'. Even when it first came out the image was pretty soft (most of the people I knew who first bought it used it primarily as a SD replacement for their DVX-100's), which is why the EX1 seemed so amazing when it was introduced and seemed to have about 4 times the sharpness of the HVX200 because it used full HD 1920x1080 chips.

[Daniel Epstein]

Well a half stop with less noise at normal settings is a lot. The F3 will look even better compared to the 200 when you boost the gain of each. Also at the same F Stop the DOF characteristics of the F3 will be very different than the 200.

[Matthias Sachal]

Both Cameras were set to 1920/25P. On the F3 I use Nikon AI Lenses on which I used the same F Stop as the HVX200. I mean the 18dB low noise is great but for me it feels like the F3 should be more sensitive all along, not just with gain enabled. The thing about the gamma curves is also right, but can it be such a difference, that it makes both cameras seem equally sensitive?

Here is a Youtube Link with the F3 using a Nikon 50mm 1.4, and a Nikon 18-105mm 3.5 zoom. And the HVX with Stock Lens with adjusted F Stops of course.

http://www.youtube.com/watch?v=EpNmNQ2kwNg

[Brian Drysdale]

You can easily lose a stop using different gamma sets ups, I've noticed the change on a number of cameras..

[David Heath]

Originally Posted by Matthias Sachal

I mean the 18dB low noise is great but for me it feels like the F3 should be more sensitive all along, not just with gain enabled.

I disagree. Most of the time, the native ISO rating is more than adequate - if it was any greater all it would mean is that more ND would be needed. Hence it's sensible to use the much greater inherent sensitivity to normally give a noise level and/or improved gamma etc abilities.

And when the light goes down, and you need the higher ISO rating, that's when you put the gain in.

My own feeling is that a far better (simple) test of two cameras sensitivity is to take them to a dark place (outside my house, streetlit at night is good - and repeatable), and via iris and gain get an exposed picture. I expect any camera to be noisy - but doing that really shows up the differences. I'd expect the F3 to be more than usable uunder such conditions, I wouldn't expect the HVX200 to be so. You may like to do just that test and let us see the results.....?

[Matthias Sachal]

The thing is that I probably expected too much from the F3 by seeing all those awesome low light tests. I once did a comparison of some ENG cameras (HDCAM, XDCAM422, HPX500) and the HVX200. The results were the 2/3" HPX500 was surprisingly the most light sensitive Camera. It was one F Stop above the HVX like the F3. I just expected the F3 with the even bigger Sensor to top that.

[Steve Kalle]

Originally Posted by Dave Sperling

Were both cameras set to 50i?
Are you using one of the Sony prime lenses that come with the F3K package? If not, (or if you're using a zoom) the f/stop on the lens (which should be accurate for computing DOF, but not for computing exposure) may be significantly different than the t/stop of the light actually getting through the lens. That's why cine lenses come with a T/stop, which in the case of a zoom is often a half stop slower than the f/stop (my recollection from the old film lenses I had that displayed both f/ and t/ stops.

Dave is totally spot on. The F-stop on SLR lenses is almost always different from the T stop, which is the actual amount of light. The F stop is usually 1/2 stop to almost 1 full stop lower than the T stop (although sometimes higher with super sensitive sensors like the Nikon D3). This is why Cine lenses always use the T stop. The best site for these measurements is DxO Mark where they show the actual T stop compared to the lens' rated F stop. For example, the $1800 Sony Alpha Zeiss 16-35 with a 2.8 F stop has a T stop of 3.3 at 16mm on an A900 camera. Also, the camera's sensor can slightly alter the actual T stop being transmitted. For example, the Sony A380 shows a T stop of 3.5 with this lens. I believe the main reason for the different T stop is the size of the pixels on the sensor.

[Leonard Levy]

Originally Posted by Matthias Sachal

of course theres less noise but is that what everyone means when they talk about the sensitive F3 Sensor

Yes that's exactly what they are talking about. Its not that its more sensitive in ordinary light at 0DB gain, That would probably be counterproductive as you would need to shoot with ND all the time. The point is you can shot at 12DB with no perceptible noise and 18DB with hardly any. That's where the sensitivity comes from.

Compare to your HVX in candlelight and you'll see an huge difference.

[Matthias Sachal]

Aha. Thanks Guys I learned a lot. I just expected the Picture to be much brighter because that was the case when I had the chance to compare the ENG Camcorders against the HVX a few month ago. But if its mainly just the usable gain, thats cool with me. Just wanted to make sure I didnt miss something.

[David Heath]

Originally Posted by Matthias Sachal

But if its mainly just the usable gain, thats cool with me. Just wanted to make sure I didnt miss something.

Well, it's more than just that - being able to use higher gain with the same or less noise obviously means better low light performance.

But when there's enough light that extra gain isn't needed, it has a lower noise figure. You gain either way.

[Brian Drysdale]

Originally Posted by Steve Kalle

Dave is totally spot on. The F-stop on SLR lenses is almost always different from the T stop, which is the actual amount of light. The F stop is usually 1/2 stop to almost 1 full stop lower than the T stop (although sometimes higher with super sensitive sensors like the Nikon D3). This is why

It's not usually that large a difference with multi coated modern lenses, but certainly 1960s zooms can have almost a stop difference. An example would be the old 16mm Angenieux 9.5-95mm, which was f2.2, with a T stop of T2.8, compared to the the later Zeiss 10-100mm zoom with f2.8 but a T stop of T3 in the late models (T3.1 in the earlier). Needless to say the Zeiss was also a much more compact design.

[Timur Civan]

You sure the F3 wasnt on -3 gain @ "low" setting?

The F3 sits around 500ISO @ -3dB approximately a half stop faster than the HVX's 320 ISO. When they are both stopped to F3.5, an approximate 1/2 stop difference sounds about right.