XF300 vs. XH-A1 Low Light/Gain/Noise

[Michael Horn]

Hi Everyone,
I thought I would post this short video I made comparing the XH-A1 to the XF300. This was shot with the default, out of the box settings on both cameras with f1.6, 1/30 shutter, no color presets, no grading in post. I hope this can be helpful to some in determining picture quality and low light and gain performance.

XF300 vs. XH-A1 Lowlight/Gain Test on Vimeo

[Glen Vandermolen]

Wow, I'm shocked!
I thoght the 300 would beat it hands down in light sensitivity, being it uses CMOS chips.

[Noa Put]

I"m surprised as well, the xh-a1 (which I currently own) has never been know for it's low light sensitivity. It does look thought that the xf-300 is a lot cleaner at +12db gain, at that setting my xh-a1 looks plain ugly, the grain then looks like a can of worms.

[Michael Horn]

I was very surprised to see these results as well. I think I may try some outdoor shots at night and see what I get next week. I'll be sure to post the results.

[Noa Put]

The thing with internet renders is that we never get to see the actual difference, +12db gain with the xh-a1 looks more or less ok in your movie but I"m sure that in reality when comparing native on a full hd screen it can look a lot worse.

We might get a wrong impression now; maybe the xf300 can look better at 15db gain compared to the xh-a1 at 6db gain? Only if you'd post native resolution pics, only then we could see the real differences.

[Glen Vandermolen]

I hate to ask, Michael, but..

You did have the ND filters off, correct? I know I've left mine on my 305 before, and was scratching my head wondering why an interior scene was so dark.

[Michael Horn]

Glen,
I thought that at first as well after seeing the initial footage, but all ND filters were off. I'll try to get some native resolution pics up soon. When I view the footage on my Full HD monitor the XF300 comes out MUCH cleaner than the XH-A1. I'll also make another video with correctly lit scenes and add gain in so we can see the effect of the noise at different gain levels in a more realistic situation.

[Noa Put]

I'll try to get some native resolution pics up soon.

That would be great! thanx

When I view the footage on my Full HD monitor the XF300 comes out MUCH cleaner than the XH-A1.

exactly what I mean, the vimeo render doesn't do the xf300 footage any justice, based on the footage in vimeo alone I would never considering that big investment to go from a xh-a1 to a xf300, in this way the xh-a1 looks almost as good and that's something I would not expect.

[Josh Dahlberg]

It may be worth noting that the XF takes a big hit shooting 1080 progressive. Switch to interlaced or 720 progressive and it gains a full stop.

If I know output is going to be dvd or web, and light is an issue, I've been shooting 720p - it makes a big difference and the images look gorgeous, albeit with a resolution loss.

[Philip Lipetz]

... And then what is the purpose of having the cam if you have to turn off it's main advantages to work in delivery formats that are less than optimal!

[Michael Horn]

Here is a link to the original .mov file before compression to web:

http://www.meridenmusic.com/video/XF...A1LowLight.mov

Just to warn you it's over 600MB, so it may take a bit to download, but you will be able to see the real life results. (right click to download)

[Glen Vandermolen]

I think the more pixels you squeeze into a chip, the less light sensitive it becomes. Each pixel becomes smaller to fit into the chip. SD cameras, with their smaller pixel count, usually had better performance in lower light then HD cams.
That might be the case here. The XF has over 2 million pixels, the XH around 1.5 million.
I know when I used an HVX200a, it always had pretty good low light capabilities, at least next to the JVC HD200 I had at the time. The HVX uses SD chips uprezzed to HD, whereas the HD200 has true HD chips, albeit only 720 capable.

[Dave Partington]

I've had an XF300 on evaluation for the last couple of days and my two reactions are :

• WOW! This camera is awesome - in good light

• Uggh! This camera is not so good in less than good light

Even at 6db I found the noise level to be less than satisfactory, and even with some neat video processing it was still less than desirable. I guess it's asking a lot from a 1920x1080 1/3" chip.

[Josh Dahlberg]

Originally Posted by Philip Lipetz

... And then what is the purpose of having the cam if you have to turn off it's main advantages to work in delivery formats that are less than optimal!

For the work I do, 90% of the time I can control the light, and there's plenty of it, so it's a non-issue.

However, it's good to know you've got an extra stop when needed, albeit with a res tradeoff; this isn't really an issue if the project will be delivered on DVD or web.

The XF is certainly no low light king: my old Sony Z5 beat it handily in this respect (it was also a step up from the XHA1). But low light isn't everything - the XF has numerous advantages over the XHA1, this isn't a key differential.

[Tim Bakland]

Can't thank you enough for posting this helpful clip.

Must say, though, and this might be an unpopular view:

The full res video (thanks for linking that, too) only confirmed my feeling when I saw the Vimeo version. I actually find that the XF300, though a stop darker, is giving the superior image in all cases here. Yes, the A1 is brighter and, at quick glance and in the smaller vimeo window, is superficially the one that jumps out. But, when blown up and seen in the full res version, it is confirmed to me that the A1 is far grainier and far inferior in terms of detail and color quality.

My 2c. Thanks again for posting.