Originally Posted by Spearing
After doing further tests, I have found the Sony CX series to be better in low light conditions than the Panny. During my first test, both cams were set at maximum 12x optical zoom in very dark conditions and the Panny was able to focus on a chair when zoomed in all the way while the Sony struggled to keep focus. Other than the focus issue, Sony's picture is brighter and has more natural looking colors indoors while the TM has a reddish/yellow tint to it, which bothers me. As others have said, both cameras have their pros and cons.
Interesting. I've read the definitions of lux (1 lumen per square meter) but suspect I'm rarely going to do what some people are listing as true "low lux" filming - 3-8 lux, say. I think I'm mostly doing 20-80 lux with the indoor filming, and very little outdoor filming though I'll be doing some at DisneyWorld next year (fireworks and the Fantasmic show).
I went to film a model train layout at Brunswick, Maryland yesterday, and also planned on filming at least a live freight train or two and the Amtrak "Capitol Limited" at Harper's Ferry. All went off as planned except it's now late enough in the year that the Amtrak train came by at what most people would call twilight or night already. I hadn't filmed any train at that low light level before.
1. Model train layout: I took film and stills of this in April 2007 with a Sony HC7 (mini-DV, 1440 instead of "full HD" but still HD). The low-light parts of the layout came out very grainy in the video and not sharp at all. The stills had to use flash or they were useless, and the flash was too bright for the subject. Fun but not good looking. By contrast, the CX500V video looked pretty much noise-free wherever it was on the layout, though I would have preferred it a little sharper sometimes. This was an HO layout so I suspect that's partly due to overall limitations in sharpness at the light level and partly due to focus never quite being perfect on small moving trains. But the video was definitely still HD quality, and only a notch below professional video I've seen of model trains which I suspect was taken in brightly lit rooms, at least for the filming. I may put a long clip together as kind of a "fish aquarium" relaxing video (running trains can be relaxing). On the stills side, the CX was much better than the HC7. Overall, though, the low flash on the CX was still overwhelming with the kinds of materials and colors involved, and the best stills would probably come from frames of the video as its normal look was halfway in-between the stills with and without flash.
2. Freight train in normal light - looked marginally better than the CX12 exposure-wise and detail-wise, but the sensor is oversampling against the HD spec anyway so I'd have to compare these more closely to say what differed. The 500 had a clear edge but not a massive one.
3. Harper's Ferry at twilight and then full dusk: this video surprised me. First, with the sun down behind the hills but not set, the video looked really good. Then, after the train went by and before full dusk, the video was noticeably brighter in the LCD than the scene was to my naked eye, and this remained true when I saw the actual footage. There's a point there where the 500 is "seeing" more light than you are and retaining colors too. This video was marginally noisier than the twilight video but still clearly HD. I did not expect any cam to perform like this, but maybe it's typical of all of them except for the low noise.
4. The train came through in "late twilight" and all the lamps and streetlights and house lights were on. Harper's Ferry in the historic district is nowhere near as bright as a normal modern town. But the cam picked up more than enough light from what was left from the sundown and the electric lights - the video almost looks like it was shot in normal daylight but not bright sun. The reflections off the side of the silver Amtrak cars are very nice. The only negative here was that the four headlights blended into one ball of light until they were about 50 yards away - possibly more so than with the CX12 footage from last year where the train came by in daylight and also was on the next track over. Anyway, the 500 looks to take point electric light sources and you see those as round balls of light if the ambient light is much lower. Again, low-light or not, the noise was very low and the video very usable.
5. I hung around 15 minutes to see if a freight would come by - no luck. By then it was full dark. I pointed the cam at the cliffs across the river and could see some rock definition in the LCD, certainly more than the dark mass I could see by eye. This wasn't useful video, though. I then flipped on the "low lux" setting for the first time and a surprising amount of detail on the rock face was visible (200 yards away). The picture was quite grainy, of course - the only light came from the lights near me and two small red train signal lights on the far end of the railroad bridge nearer the cliff. I didn't expect the low lux mode to show any detail of the rock face at all, so again I was surprised at how much better the cam could "see" vs my own eyes at that point.
Hmmm...that was long. Sorry. One last comment, again about something advertised but not taken super seriously. I took a tripod with me but decided never to use it to check the stabilization improvements. Either I have become a master videographer since 2007 or the stabilization on the 500 is every bit as good as other people have reported. It was phenomenal by comparison with the 2007 model train footage. In fact, for most of the footage for the whole day, I suspect non-filming people would not realize the cam wasn't on a tripod at any point. In 2007, my film of moving model trains jerked around in minor ways constantly. I was a little horrified watching it tonight. Then I switched over to the new footage and was much relieved. It was very smooth and I was able to pivot to follow trains and even walk after them in a few cases. Similarly, filming the actual freight and Amtrak trains, it was very close to tripod-quality. For the latter, I followed the engine on foot for about 25 yards after it went by and that footage stabilized beautifully, too. If I can do this as a rank amateur, I'd love to see what someone with top-notch skills could do off-tripod with these cams.
Bottom line: the improvements in the last 2-3 years from the HC7 to the CX500V are significant. The improvement from the CX12 to the CX500V in low-light quality and stabilization is signficant. I'm not sure the low lux mode has benefited as much as the filming in low light with regular settings, so I'm not really commenting on that here. I can't compare vendor cams to point out which is best, but I can say my opinion is that Sony has improved their cams I've bought a great deal in the last three years. I agree with other peoples' comments about the manual dial and no zebra and so forth as sometimes steps backwards. But overall, I'm getting improvements where I need them the most from what I can tell.
Excluding the optics, we're talking computers here now, so the race is on between vendors to pack in effective new features via hardware and software improvements. That should benefit all of us. If the video looks as different 3 years from now as it does versus 3 years ago, that could get very interesting.