GH1 and T2i which is the one, no flame here

[Samuel H]

How are you guys getting good focused images if you can't monitor it externally? Also isn't the GH-1 screen lower resolution compared to the T2i screen?

I understand the histogram view is on the GH-1 but is it still hard to get great focused images just with that?

I also wondered that, I guess the answer is on the viewfinder (which the GF1 lacks)
GH1 LCD screen: 460.000 dots (480x320 ??)
GH1 EVF: 1.440.000 dots (900x540 ??)
T2i LDC screen: 1.040.000 dots (680x510 ??)

(sorry, I hadn't seen the previous reply)

 

[zcream]

One thing I have not understood.
7D/T2i record in mpeg4 at 48mbps
Since MJPEG does not allow 25P24P, the comparison is AVCHD at 30mbps (and despite the 60,62, 70 settings I have NEVER seen it go higher).
If both are interframe, then the GH1 pic should have more artifacts/ However I am not seeing this in comparison between T2i vs GH1.
Unless the 7D is intraframe.
The compression really makes no sense.

 

[Adventsam]

Couple of points

Couple of points

One thing I have not understood.
7D/T2i record in mpeg4 at 48mbps
Its all AVC H264 though
Since MJPEG does not allow 25P24P, the comparison is AVCHD at 30mbps (and despite the 60,62, 70 settings I have NEVER seen it go higher).
I am regularly getting 35mbit+avg, 40+mb spikes in 50i, 25p will be higher IMO, but avc is vbr I think as std. What Canon did is a custom specific implementation. CBR of a from in GH1 would be good, but am not sure its possible, maybe Vitaliy can work it out.
If both are interframe, then the GH1 pic should have more artifacts/ However I am not seeing this in comparison between T2i vs GH1.
Not sure, dont have a 7d
Unless the 7D is intraframe.
The compression really makes no sense.

Cheers

 

[zcream]

Hi Adventsam.
So the Canons use a CBR AVC H.264 fixed rate interframe.
The GH1 used a VBR AVCHD that is maximally in the 30s with spikes in the 40s.
Its still more compression than the Canons.

 

[Adventsam]

Possibly and may explain

Possibly and may explain

why many prefer low light Canon footage, but am no expert here, would need to see side-side tests. Not sure 100% about Canon CBR, but I think AFAIK is case.

 

[Barry_Green]

Its still more compression than the Canons.

Not necessarily. It is the same compression family, but do we know that they use the same compression tools?

You cannot just compare them and say "well, they're both h.264, so the only difference is bitrate." Because AVCCAM AVCHD is h.264, but at 21mbps PH mode, it's dramatically superior to what the Canons do in 50mbps.

h.264 has many, many tools that can be employed, and the codec implementer has the freedom to choose what they will or won't use.

The only way you could fairly compare GH1 and 7D, based on bitrate alone, is if you knew for a fact that they were both using all the same toolsets. And I don't think such information has ever been released or verified. The debate about VBR vs. CBR is just one of many potential explanations that may or may not be accurate.

What we're missing here is a relative "efficiency" rating for how the codec is performing. The HA mode in AVCCAM is very efficient, squeezing the full image into 17mbps. The FHD mode in the stock GH1 is very inefficient, producing an image in 17mbps that can't hold a candle to the AVCCAM version. The AVCCAM HA mode is hugely massively better than the GH1's FHD version, even though they're both h.264 at 17 megabits.

So -- just because the GH1 and Canon are running at more-comparable megabits now, doesn't mean that the compression is equivalent between them.