Interesting Low Light Tests at DVinfo.net

Miguel Lombana · August 24th, 2004, 12:37 PM

Ok so what do you do when a 4 month old doesn't want to sleep? You video her toy area with your camera to see what type of low light performance it has, ok I was really bored.

Anyway I wanted to test some stuff and find the best combination for presets, options, light levels etc and took advantage of the fact that I was up late with the baby so here goes, and there is one very interesting finding.

I shot these scenes in a 15' x 20' room where I was at the extreme opposite side of a 60watt diffused lamp pointing at the cieling so it was fairly night time, not "theatre" dark but typical late night at home dark.

What I found to be intesting is the results in this light environment when in "Green Box" mode. From my very unscientific testing, it appeared to me that the best results were from this mode, the noise was the lowest with the picture being very usable with and without the tiny SIMO 3watt light that I used.

White balance was auto, settings used are all on the images, the JPG files were downsized to a quality of 8 in photoshop from the original BMP that Premier exported and again the results are to say the least, interesting.

The pics are at:
http://www2.dj-miguel.com:8000/modul...nails&album=25

*** Forgot to mention, click on the Date + flag to sort the pics out the way they were intended on being viewed.

Regards,
MIGUEL

Thomas Smith Jr · August 24th, 2004, 07:43 PM

First off i'd like to say thank you for taking your time to do tests

But these are JPEGS. When posting screen grabs like these, it's really important that you don't use compression at all on the full size (720x480) images. Since your full size images are like 75 kb, I am assuming you compressed them a little for the web. JPEG compression leaves artifacts and will leave a negative impact on what is supposed to be a true representation of that particular light setup and camera setting and ultimately make an image look worse than it actually is. It's important that you don't do that.

These grabs have succesfully re-affirmed my hatred for gain. If I saw these grabs before I bought my Gl2, I wouldn't have bought a Gl2.

Miguel Lombana · August 24th, 2004, 08:00 PM

<<<-- Originally posted by Thomas Smith Jr : First off i'd like to say thank you for taking your time to do tests

But these are JPEGS. When posting screen grabs like these, it's really important that you don't use compression at all on the full size (720x480) images. Since your full size images are like 75 kb, I am assuming you compressed them a little for the web. JPEG compression leaves artifacts and will leave a negative impact on what is supposed to be a true representation of that particular light setup and camera setting and ultimately make an image look worse than it actually is. It's important that you don't do that.

These grabs have succesfully re-affirmed my hatred for gain. If I saw these grabs before I bought my Gl2, I wouldn't have bought a Gl2. -->>>

The original output was from Premier which dumps to a bitmap file, the size was about 1meg but for the web I did have to convert to JPG and compress, I opted to do this due to my lack of bandwidth, I can however re-cut a couple of these images with no compression in a differnt format and post them to a static html page and not the photo gallery I use on the PHP site.

What I'd really like to do (which will likely break my heart given that I think I know what the results are going to be) is borrow a VX2100 and do a side by side.

Bottom line it was really interesting that for this very non-scientific test, running in green box produced an image that was more palatable than auto or manual. On the flipside, the use of just the little 3w lite at 3 feet made such a huge difference for some of these shots such as the 12db 1/30th. I would almost opt to use this for some close in wedding work where I didn't want to blind someone with the 25w + Frezzi.

Overall, it was a quick and easy little test, again non-scientific but just gave me something to do.

Miguel

Bob Harotunian · August 25th, 2004, 07:16 AM

There's nothing new here folks. After all, this camera has been around now since August 2002. We should all know by now that the GL2 is a great overall camera except for low-light conditions.
Unless you're often shooting in low-light (like I am), enjoy the camera for what it delivers and remember what you paid for it vs VXs and 170s.

The only thing we should be hoping for now is that Canon wakes up and releases a GLx with truly competitive pro features. None of this prosumer crappola, but 1/3 CCDs, XLRs and 3 lux would be a nice start...sort of a handheld Xl2. Based on our experience using GL2s professionally, I would gladly pay considerably more for such an upgrade.
Bob

August 24th, 2004, 12:37 PM	#1
Miguel Lombana Major Player Join Date: Mar 2004 Location: Litchfield Park, AZ (W/of Phoenix) Posts: 502	Interesting Low Light Tests Ok so what do you do when a 4 month old doesn't want to sleep? You video her toy area with your camera to see what type of low light performance it has, ok I was really bored. Anyway I wanted to test some stuff and find the best combination for presets, options, light levels etc and took advantage of the fact that I was up late with the baby so here goes, and there is one very interesting finding. I shot these scenes in a 15' x 20' room where I was at the extreme opposite side of a 60watt diffused lamp pointing at the cieling so it was fairly night time, not "theatre" dark but typical late night at home dark. What I found to be intesting is the results in this light environment when in "Green Box" mode. From my very unscientific testing, it appeared to me that the best results were from this mode, the noise was the lowest with the picture being very usable with and without the tiny SIMO 3watt light that I used. White balance was auto, settings used are all on the images, the JPG files were downsized to a quality of 8 in photoshop from the original BMP that Premier exported and again the results are to say the least, interesting. The pics are at: http://www2.dj-miguel.com:8000/modul...nails&album=25 *** Forgot to mention, click on the Date + flag to sort the pics out the way they were intended on being viewed. Regards, MIGUEL __________________ Miguel Lombana http://www.miguellombana.com & http://www.phoenixhamradio.com

August 24th, 2004, 08:00 PM	#3
Miguel Lombana Major Player Join Date: Mar 2004 Location: Litchfield Park, AZ (W/of Phoenix) Posts: 502	<<<-- Originally posted by Thomas Smith Jr : First off i'd like to say thank you for taking your time to do tests But these are JPEGS. When posting screen grabs like these, it's really important that you don't use compression at all on the full size (720x480) images. Since your full size images are like 75 kb, I am assuming you compressed them a little for the web. JPEG compression leaves artifacts and will leave a negative impact on what is supposed to be a true representation of that particular light setup and camera setting and ultimately make an image look worse than it actually is. It's important that you don't do that. These grabs have succesfully re-affirmed my hatred for gain. If I saw these grabs before I bought my Gl2, I wouldn't have bought a Gl2. -->>> The original output was from Premier which dumps to a bitmap file, the size was about 1meg but for the web I did have to convert to JPG and compress, I opted to do this due to my lack of bandwidth, I can however re-cut a couple of these images with no compression in a differnt format and post them to a static html page and not the photo gallery I use on the PHP site. What I'd really like to do (which will likely break my heart given that I think I know what the results are going to be) is borrow a VX2100 and do a side by side. Bottom line it was really interesting that for this very non-scientific test, running in green box produced an image that was more palatable than auto or manual. On the flipside, the use of just the little 3w lite at 3 feet made such a huge difference for some of these shots such as the 12db 1/30th. I would almost opt to use this for some close in wedding work where I didn't want to blind someone with the 25w + Frezzi. Overall, it was a quick and easy little test, again non-scientific but just gave me something to do. Miguel __________________ Miguel Lombana http://www.miguellombana.com & http://www.phoenixhamradio.com

August 24th, 2004, 07:43 PM	#2
Thomas Smith Jr Regular Crew Join Date: Jul 2004 Location: Oxnard, CA Posts: 26	First off i'd like to say thank you for taking your time to do tests But these are JPEGS. When posting screen grabs like these, it's really important that you don't use compression at all on the full size (720x480) images. Since your full size images are like 75 kb, I am assuming you compressed them a little for the web. JPEG compression leaves artifacts and will leave a negative impact on what is supposed to be a true representation of that particular light setup and camera setting and ultimately make an image look worse than it actually is. It's important that you don't do that. These grabs have succesfully re-affirmed my hatred for gain. If I saw these grabs before I bought my Gl2, I wouldn't have bought a Gl2.

August 25th, 2004, 07:16 AM	#4
Bob Harotunian Major Player Join Date: Mar 2003 Location: Somerville, MA Posts: 360	There's nothing new here folks. After all, this camera has been around now since August 2002. We should all know by now that the GL2 is a great overall camera except for low-light conditions. Unless you're often shooting in low-light (like I am), enjoy the camera for what it delivers and remember what you paid for it vs VXs and 170s. The only thing we should be hoping for now is that Canon wakes up and releases a GLx with truly competitive pro features. None of this prosumer crappola, but 1/3 CCDs, XLRs and 3 lux would be a nice start...sort of a handheld Xl2. Based on our experience using GL2s professionally, I would gladly pay considerably more for such an upgrade. Bob