550D/GH13 comparision shot

Im sure Kholi is exactly right.

I can't compare cameras but I've shot some scenes with very little light and they have turned out extremely well with the T2i.

This is shooting with much lesss than one normally would. Not unlike that shot of DiCaprio holding a match as the light source illuminating his face. I believe that was on a RED.

I haven't gone that far though. :)

MAH

RED MX, and none of these DSLRs hold a match to the image.

I don't get the obsession with "lowlight" on these things. It all looks terrible to me. If I'm gonna shoot under street lamps it'll be MX.

I know for sure that I'll shoot FAR LESS "lowlight" than actual lit or well exposed material. I surely wouldn't be buying a camera based on conditions that I would rarely shoot in.
 
Kholi, I'm not saying it can hold a match to a RED. Just saying that it can be lit with very little and still look pretty damn good. I think all of the cameras have improved in this way.

It's not an obsession. It's just handy in many circumstances. Most of which you pointed out earlier.

Bloom has posted about it many times and it's a plus to some people. It doesn't need to be a plus for you. No one is forcing it on you. I'm speaking for myself, as I said earlier.

Cheers,

MAH
 
RED MX, and none of these DSLRs hold a match to the image.

I don't get the obsession with "lowlight" on these things. It all looks terrible to me. If I'm gonna shoot under street lamps it'll be MX.

I know for sure that I'll shoot FAR LESS "lowlight" than actual lit or well exposed material. I surely wouldn't be buying a camera based on conditions that I would rarely shoot in.

Not pertaining to DSLRs in particular, but cameras that can shoot in low light mean you can use less powerful lights and not require so many lights. You still need lights, absolutely, but it opens up the door to be more creative.

In most of the shoots I have been on with the 7D, like recently a basement, it was really nice to work in such a dark location with a few work lights and really have a lot of freedom.

When we shot with other cameras that required A LOT of light to get a decent noiseless image, we spent a lot of time fighting the light positions to get enough light on ALL parts of the shoots.

This is what I love about the MX sensor as well, it works so well in low light that I look forward to this trend continuing and newer lighting products coming out that run cooler, use less power, and can be hidden into a scene easier.
 
This is what I love about the MX sensor as well, it works so well in low light that I look forward to this trend continuing and newer lighting products coming out that run cooler, use less power, and can be hidden into a scene easier.
Amen 2 that!!
 
RED MX, and none of these DSLRs hold a match to the image.

I don't get the obsession with "lowlight" on these things. It all looks terrible to me. If I'm gonna shoot under street lamps it'll be MX.

I know for sure that I'll shoot FAR LESS "lowlight" than actual lit or well exposed material. I surely wouldn't be buying a camera based on conditions that I would rarely shoot in.

:happy: who needs a RED for Lowlight ??? me not



http://vimeo.com/12720689
 
Last edited:
Kindly point me in the direction of just one of these tests, and you'll have made my evening
ISOs are a standard. now with that in mind, the question is how noisy is gh1 at iso 800+?
the 7d tops at iso 1200 for acceptable noise for me so,
if you find that answer out then you'll know the answer.
 
I'm about to shoot a feature in late august and I've been thinking of buying a GH1 just to shoot my wides/deep focus with and use my t2i for all the medium/close stuff. Since all my testing is in line with Barry's thoughts about the t2i/7d being bad under deep focus shooting conditions (and most of my feature will be shot in the woods durring the day time/which means lots of exposure range with a need for detail both the highlights and lowlights). My thought is that even if I just buy a GH1 for these shots and check the frame on an HD monitor before hand because the shot will most likely be locked off and extreamly wide I can get away with not needing the HDMI out during the shot. Anyone have any ideas on how the two camera's footage will intercut?
 
I bought a T2i in May, upgrading from an HV20 and haven't looked back. The 24p just "snaps" better than my HV20 ever did. I notice less rolling shutter than I did with the HV20. The motion of the T2i when set at 24p (shutter 50) reminds me of my original DVX100B which brings a big smile to my face.

These comparisons have been quite interesting to study. Is it possible that the compression codec is to blame for some of the murky details in the T2i shots? It seems I always notice this in areas of similar color and texture (grass, tress in distance, sand, etc?). The question: is this codec related?

I happen to like the contrast of the T2i shots. It adds a depth to the overall image (the ducks in particular). on the other hand, the detail in the HMC40 and GH13 shots is unquestionably great. Once again, is this strictly codec related? I understand somewhat about line skipping and how the sensors work. Somebody just has to experiment tweaking that internal H264 codec in the Canons. Till then, I'm happy with my purchase.

I think we all owe Barry a round of applause and thanks for doing this test. The level of professionalism that he used to ensure that this test was foolproof from the beginning is certainly to be commended and respected.
 
Last edited:
I'm about to shoot a feature in late august and I've been thinking of buying a GH1 just to shoot my wides/deep focus with and use my t2i for all the medium/close stuff. Since all my testing is in line with Barry's thoughts about the t2i/7d being bad under deep focus shooting conditions (and most of my feature will be shot in the woods durring the day time/which means lots of exposure range with a need for detail both the highlights and lowlights). My thought is that even if I just buy a GH1 for these shots and check the frame on an HD monitor before hand because the shot will most likely be locked off and extreamly wide I can get away with not needing the HDMI out during the shot. Anyone have any ideas on how the two camera's footage will intercut?
I don't know how well they'll intercut, but I would imagine you'll get them to match pretty well in post...
 
These comparisons have been quite interesting to study. Is it possible that the compression codec is to blame for some of the murky details in the T2i shots? It seems I always notice this in areas of similar color and texture (grass, tress in distance, sand, etc?). The question: is this codec related?
Almost certainly. I say this because prior to the GH13 hack, the GH1's shots looked just like the 7D's, only worse. Nothing changed in the hardware, only the codec, yet all of a sudden the GH13 is just way sharper/clearer/crisper in those same scenarios, and the only improvement is in the codec.

There may be other complicating factors in the 7D/t2i (such as the color moire) that may be aggravating the situation, but we can't know that without isolating the variables. It could be the codec, or it could be an issue introduced by the way the sensor line-skips/pixel-bins for video, we don't know. If someone has an external HDMI recorder, that would be a way to test.

I happen to like the contrast of the T2i shots. It adds a depth to the overall image (the ducks in particular).
Contrast is a huge influence on perceived image quality, and one of the reasons why the HVX200 always held its own or excelled in comparison shots against sharper cameras. There have been studies done that say that contrast is a far more important factor in how we judge the appeal of images than resolution is.

But, be aware that no effort at all was made to enhance the contrast in my shots. It was left as flat as possible. You could probably tweak any of these to get the type of contrast you want.

I think we all owe Barry a round of applause and thanks for doing this test. The level of professionalism that he used to ensure that this test was foolproof from the beginning is certainly to be commended and respected.
You're very welcome and thanks to all who participated and for taking it exactly as intended and being so professional and reasonable about it, without any flames or emotions injected. :)
 
I don't know how well they'll intercut, but I would imagine you'll get them to match pretty well in post...

Thanks Barry. I'm going to run some tests, I need to nail this down asap so I can walk into this project without any worries. I've been noticing color shifts with my t2i at wider focal lengths and it's driving me crazy...
 
Thanks Barry. I'm going to run some tests, I need to nail this down asap so I can walk into this project without any worries. I've been noticing color shifts with my t2i at wider focal lengths and it's driving me crazy...


What you describe is exactly what I meant: I'd be worried about getting a camera that does MOST of what I need to do, not small parts of what I can get other cameras for.

If a lot of your material's out in the wild with a lot of detail, then you have to consider that.

I would definitely find someone with a hacked Camera and just go out and shoot, see what you like for yourself. But, so far in what you've described I would definitely say GH13 or even an actual video camera.
 
What you describe is exactly what I meant: I'd be worried about getting a camera that does MOST of what I need to do, not small parts of what I can get other cameras for.

If a lot of your material's out in the wild with a lot of detail, then you have to consider that.

I would definitely find someone with a hacked Camera and just go out and shoot, see what you like for yourself. But, so far in what you've described I would definitely say GH13 or even an actual video camera.

I am in a very strange situation right now, this feature coming up is in many ways "my shot" it will be my third feature and I feel that if everything isn't perfect then I will have blown a huge op for me. I am considdering the t2i for close work and a GH13 for wides. I have also been thinking of shooting the entire thing on a Red but I like having more than one cam on set and I can't afford to have more than one red...
 
I think if I were in your shoes (I've never made a feature so take this with a grain of salt) I would go with the Red if it's an option.

One Red is better than no Red. :)

MAH
 
I am in a very strange situation right now, this feature coming up is in many ways "my shot" it will be my third feature and I feel that if everything isn't perfect then I will have blown a huge op for me. I am considdering the t2i for close work and a GH13 for wides. I have also been thinking of shooting the entire thing on a Red but I like having more than one cam on set and I can't afford to have more than one red...

I'll tell you right now: GH13 will slice with RED no problem. On a large screen you'd be hard pressed to tell the difference, if mastering at 1080/2K.

For a small screen, T2i/7D will also cut in nicely. Our West Fest short needed pickups and we had to run light, so instead of dragging one of the REDs out and getting the anamorphics again, we used the 7D and some of my cheap old glass to match the old look of the Lomo Anamorphics.

When the shorts up, you'll be able to see which shots are 7D I think, but I don't know about a regular audience member.

I'm pretty confident that a GH13 + Hot Rod PL mount + Same glass as RED (cooke S4's please!) would be damned near indistinguishable on any large source if cared for properly.

All that to say: if this is your third feature, why skimp? You've got history, you can get a RED package for dirt cheap and reap the benefits. Any lowlight shot you can take care of with an MX sensor @ high ISO (trust me, it's unbelievable when you get it in your own hands) and you can cut it up with the DSLR.
 
I've thought of that but I'm of the mindset that more coverage and more cams (we have two t2i's and could have 2 hvx200s and one hmc150 if we wanted) make's shooting faster and cutting down on time that we have talent on set can save us ALOT... I'll stop hijacking this thread now lol...
 
I'll tell you right now: GH13 will slice with RED no problem. On a large screen you'd be hard pressed to tell the difference, if mastering at 1080/2K.

For a small screen, T2i/7D will also cut in nicely. Our West Fest short needed pickups and we had to run light, so instead of dragging one of the REDs out and getting the anamorphics again, we used the 7D and some of my cheap old glass to match the old look of the Lomo Anamorphics.

When the shorts up, you'll be able to see which shots are 7D I think, but I don't know about a regular audience member.

I'm pretty confident that a GH13 + Hot Rod PL mount + Same glass as RED (cooke S4's please!) would be damned near indistinguishable on any large source if cared for properly.

All that to say: if this is your third feature, why skimp? You've got history, you can get a RED package for dirt cheap and reap the benefits. Any lowlight shot you can take care of with an MX sensor @ high ISO (trust me, it's unbelievable when you get it in your own hands) and you can cut it up with the DSLR.

I'm looking forward to seeing your short! Man am I gonna look hard at everything....
 
Hello gang!

I just purchased a GH1 to go along my 550D. I applied firmware patch to GH1 and took both both cameras out for some testing.

GH1 is so much better it's not even funny...I used kit lenses on both cameras and zoomed so I framed approximately same area. And yes, I double-checked that focus was perfect on both cameras and used 1080p on both. This is a crop from raw frame, taken from VLC. No sharpening and no grading, just frame-grab directly to file and crop.

GH1 has zero, I repeat, zero moïre when in AVCHD-mode. Tried both on brick-walls. 550D starts color-dancing and looses detail long before Panny. Panny only blurs when bricks cannot be resolved any more. To my eye, this looks like pixel-binning.

I shot some other scenes with lot's of details and difference is even greater. We would really really need a firmware upgrade for 550D right now. 550D is 18MP, panny is 13MP. I understand that pixel-binning is harder to do on higher pixel count, but if Panny can do pixel binning in 1080p30 (30x13MP = 1,17Gpixel/sec) then Canon wouldn't be far-off if doing it in 24p (24x18MP = 1,29Gpixel/sec).

This DSLR stuff is moving way too fast... :crybaby: Of well, 550D still takes superior stills.

How about this. Remember, this is frame-grab which I had to re-compress to JPEG so I can upload it so it lost a little bit more, then it got further mangled by this site when uploaded. I even took Panny footage and extrapolated chroma from 4:2:0 to 4:2:2 and then saved as RGB raw for some grading. It holds up pretty well.

I'll try to upload some 4:2:2 RAW extrapolated grading comparison as well. Panny bitrate is jumping all over the place (i used C setting), 550D is steady 45Mbit. But on many occasions, Panny bitrate was actually lower than 550D one, despite having much more detail.

My hunch is that Canon compression is a bit easier on chroma, but it's just a hunch. Need to examine...

Which is which?

Guys... Please label, not nice to guess.

I'm a noob and I have no idea about the characteristics of each system.

Of course from the comments I'd say the clearer and with most resolution would be the GH1, but you never know.

:happy: who needs a RED for Lowlight ??? me not



http://vimeo.com/12720689

The video is a black hole with very crunchy blacks and sparse bright subjects.

Did I miss the joke?
 
For all of you considering shooting with GH13:

Firmware hack opens lot's of possibilities. But *make sure you test* before you shoot anything more than ducks. This hack is still experimental and enabling certain combination of options might make you camera stop in the middle of shoot.

For time being, if you want to shoot with GH13 "for real" I believe this is wise:

1. Stay clear from MJPEG unless you can use 720p30 footage. (24p MJPEG is not stable yet and 1080p MJPEG is just stretched 720p).
2. Use AVCHD and go trough "a","b","c" bitrate settings test until you are 100% sure it works without triggering buffer overflow.
3. Use Sandisk Class 10 SD cards.

That being said, there are couple of things you can do which cannot be done with Canons. If using MJPEG, you can use shutter speeds lower than frame rate. (if using 1/15 shutter, actual framerate will drop to 15fps but it will still get rendered into 30 frames per second, doubling each frame).
 
Last edited:
These comparisons have been quite interesting to study. Is it possible that the compression codec is to blame for some of the murky details in the T2i shots? It seems I always notice this in areas of similar color and texture (grass, tress in distance, sand, etc?). The question: is this codec related?

Maybe it's noise reduction? I have a first generation 1DS, and when I got the first generation 5D, I thought the 1DS was sharper, even though it had fewer megapixels. The 5D had far better high iso performance, so probably more noise reduction, which smoothed out some of the details. Maybe the same thing's going on with the video?

Maybe a hacker can dive into the 7D/T2i firmware and reduce noise reduction, bringing back the detail (probably at the expense of high ISO performance, however).
 
Back
Top