View Full Version : Using a monitor on the 7d


Phil Bloom
September 15th, 2009, 12:08 PM
It's so much better than a 5dmkII. I hadn't tried it out until today but it doesn't drop down to 480p and change aspect ratio when you hit record. It's still not FULL HD out but it's leaps and bounds better than with the 5dmkII

Philip Bloom Blog Archive Using a monitor on the 7d is much better than with the 5dmkII (http://philipbloom.co.uk/2009/09/15/using-a-monitor-on-the-7d-is-much-better-than-with-the-5dmkii/)

Brad McGiveron
September 15th, 2009, 12:25 PM
That's some great news Phil !!

The 7D seems to be looking better and better by the day.

Jon Fairhurst
September 15th, 2009, 12:29 PM
This is HUGE.

Not only is the 7D easier to focus than the 5D2 because of its smaller sensor, but the higher quality monitor will make live critical focus possible.

The only downside of the 7D that I can see is the lack of fast, wide primes.

Man, oh man, do I hope that the 5D2 firmware is upgraded to give 24p - and that either Canon or Magic Lantern give us higher resolution monitoring.

Hmmm. I could sell the 5D2, buy a 7D and spend the change on an HDMI monitor...

Burk Webb
September 15th, 2009, 12:33 PM
That is great news!!!!!

That is the main thing that bugged be about the 5D. Do you know the resolution it is sending out when rolling?

Thanks a bunch for all the reports on this camera Mr. Bloom, I am turning inside out waiting to get my hands on one of these things!

Evan Donn
September 15th, 2009, 01:29 PM
The only downside of the 7D that I can see is the lack of fast, wide primes.

Exactly, and until I had the 5D I never would have imagined how important that would be to me - but now the thing that keeps me from considering the 7D is the impact it would have on wide angles!

Man, oh man, do I hope that the 5D2 firmware is upgraded to give 24p - and that either Canon or Magic Lantern give us higher resolution monitoring.

No real knowledge here, just speculation, but I'd guess the 7D is able to do the high-res monitoring while recording due to the second DIGIC, so I'm not too confident we'll ever see it on the 5D.

Matt Newcomb
September 15th, 2009, 02:32 PM
If it's not full HD and it's not 480p, what is it exactly outputting as then? 720p?

Jenn Kramer
September 15th, 2009, 02:59 PM
It looks like it's inset, might be the base resolution of the preview lcd scaled to fit 1080.

Phil's latest video: TwitVid - @philipbloom quick demo of 7d video monitoring, filmed at WEVA (http://www.twitvid.com/9BCE7)

Khoi Pham
September 15th, 2009, 03:04 PM
It said 1080i 60hz on the top left corner of that monitor when it received the signal so it must be output full 1080i, it is at the end of the video.

Jenn Kramer
September 15th, 2009, 03:11 PM
The mode changed when he switched to playback, so we know that playback is full 1080, but it didn't show what the mode was before. I wouldn't be surprised if it was also 1080, but we'll have to wait for someone with a camera to confirm.

Khoi Pham
September 15th, 2009, 03:27 PM
yeah you are right, can't tell much but as far as I can see on the web, the recording picture looks identical to the playback picture so maybe it really does output full hd during recording.

Chris Barcellos
September 15th, 2009, 03:50 PM
My guestimate:

If it is outputting full HD, the height would be 4;3 with 1080 height, and the width of the image, since it is only a 4:3 screen would be 1440 ?

That would make the image in the 16 x9 view area, in the monitor about 810 high, or 810 x 1440. Anybody ??

Tramm Hudson
September 15th, 2009, 03:58 PM
That would make the image in the 16 x9 view area, in the monitor about 810 high, or 810 x 1440. Anybody ??
My original computation from another thread (also a guess!):

However... while the output signal is 1080, the image region is 3:2 in the best case (no Canon info displayed, after calling lv_output_device( "1080full" )). This means there are side-pillars on both sides of the 16:9 1080i signal, and then soft-mattes to show what portion of the 3:2 is the 16:9 recorded image.

So the effective resolution is even less than the 1080i signal might suggest. I haven't measured it exactly, but my estimate is that we lose about 150 pixels per side-pillar, leaving us with a 1620x911 useful image on the HDMI. I'm still investigating to see if it is possible to get different ratios and to fill the entire screen.

Nick Hiltgen
September 16th, 2009, 06:41 AM
Phil when you plug in the monitor do you lose the ability to monitor on the LCD? If not do you know if you get the option to monitor via both outputs at once?

Tramm Hudson
September 16th, 2009, 09:03 PM
It's so much better than a 5dmkII. I hadn't tried it out until today but it doesn't drop down to 480p and change aspect ratio when you hit record. It's still not FULL HD out but it's leaps and bounds better than with the 5dmkII
I've just figured out how to make the 5D continue to output 1080i while recording. Here is a short demo video: Magic Lantern - 1080i while recording on Vimeo (http://vimeo.com/6618825)

Jenn Kramer
September 16th, 2009, 09:14 PM
Sweet! Nice job, Tramm!

Evan Donn
September 17th, 2009, 12:43 PM
I've just figured out how to make the 5D continue to output 1080i while recording.

Well that blows my theory about needing the second chip to do this - I wonder what the second chip does in the 7D? Do you have any way to monitor processor load or temp? I wonder if running full res will tax the system beyond some level that Canon is comfortable with.

Jon Fairhurst
September 17th, 2009, 12:54 PM
I wonder what the second chip does in the 7D?

I believe that its main function is to enable high-speed continuous shooting of 8 images per second. It's possible that it's needed for 720p60 encoding, but I'm not sure. (I doubt it.)

Two processors are definitely not needed for 1080p24, since 24p is slower than 30p. The 5D2 might have a hardwired clock that makes it impossible to upgrade to 24p with firmware alone, but it definitely doesn't need another processor to go slower.

Chris Barcellos
September 17th, 2009, 01:19 PM
I believe that its main function is to enable high-speed continuous shooting of 8 images per second. It's possible that it's needed for 720p60 encoding, but I'm not sure. (I doubt it.)

Two processors are definitely not needed for 1080p24, since 24p is slower than 30p. The 5D2 might have a hardwired clock that makes it impossible to upgrade to 24p with firmware alone, but it definitely doesn't need another processor to go slower.

This brings up a question about which I have not expertise. I remember days of overclocking motherboard chips, how about the idea of slowing the clock in the camera to get defacto 24p even though the chip thinks it running at 30p.?

Tramm Hudson
September 17th, 2009, 03:01 PM
I wonder what the second chip does in the 7D? Do you have any way to monitor processor load or temp? I wonder if running full res will tax the system beyond some level that Canon is comfortable with.
My guess is the same as Jon's -- the higher frame rate modes require a faster scan of the chip. This will also reduce jello if the sensor can be scanned in 1/80th of a second rather than 1/40th.

I'm not very confident that we'll be able to do much with a windowed read of the sensor. In LiveView mode the noise goes up dramatically while reading a small section, which makes me think that either the NR algorithm requires the entire scanline, or that reading a small section quickly requires the ISO to be bumped temporarily.

Mike Calla
September 17th, 2009, 07:35 PM
NR algorithm requires the entire scanline, or that reading a small section quickly requires the ISO to be bumped temporarily.


hahaha Tramm, i love reading your posts... I don't understand them but i love it that someone does AND is doing something about it!

On behalf of everyone, thank you!

Mike Calla
September 17th, 2009, 07:44 PM
tramm,

just like modern TVs modulate scan line frequency to improve picture quality, does the 5/7D adjust cmos scanning frequency to compensate for different shutter speeds?

i.e.: if the shutter speed is set to 1/40 will the camera scan slower than say if the shutter speed is set to 1/80?

Jon Fairhurst
September 18th, 2009, 01:19 AM
The 5D2 takes 1/40 seconds to scan the sensor, regardless of the shutter speed. This means that we get the same rolling shutter no matter what adjustments we make to the camera controls.

Unfortunately, the sensor scan time is quite unlikely to be improved by firmware. The only good news is that it forces us to get better stabilization gear, which helps us make better films. :)

Don Miller
September 19th, 2009, 04:25 PM
Do we know the scan time for the 7D?

Jon Fairhurst
September 20th, 2009, 10:11 AM
Do we know the scan time for the 7D?

Not yet.

All we need is for somebody with a prototype to film some flashes at about 1/40 shutter. Then we need to see three full-res screenshots

(1) With the entire flash captured within a frame,
(2) with some of the the flash captured in the bottom lines of a frame, and
(3) the frame immediately after (2) with the end of the flash captured at the top of the frame.

Tell us your frame rate, count the lines, and you can find out the shutter speed, the time between scans, and the time to scan the sensor.

Hopefully, it's faster than the 5D2. If Canon were to update the 5D2 to cover 7D frame rates, the two bodies would make a killer pair. (As of now, they have no common frame rates.)