digitalbug30d wrote:
well I am giving Canon some slack knowing some issues arise with higher pixel counts,I am willing to bet having dual digic 4s will take care of most of the issues
in these arguments I havent seen any talk about the processors.With 2 of them
should in theory out do a single chip camera ie the 50D,so this is why I think the
7d would be better than any single processor designed body and most likely the A/F
speed should be better,ISOs,noise ect
I would assume the number of processors does nothing to improve noise performance. Noise reduction speed would probably be improved.
Also I don't think multiple processors can improve a sensors ISO performance.
I'm ignorant on AF speed. Faster processing is probably a good thing, but do the Digics have anything to do with AF, or is there a separate processo dedicated to that?
You mention there should be improvements over the 50D. Yes. The big one is 8 FPS with an 18 MP sensor. That's a lot of data to move that is only possible with dual processors.
Does anyone have electrical schematics of thr 7D? I'd like to see those
Can someone tell me what these white and red specs are? ISO 3200 and the second is a 100% crop of underneath the SUV. I put the lens cap on and took a shot at F22. I have no dead or hot pixels. I remember reading about it but trying to find it is not easy. This is my replacement 7D and everything else is OK but I noticed this.
It seems to be more noticeable with this unit. I just found out about PS Despeckle and applied it for the white specs. Still curious as to what causes it? Here is one after Despeckle. I have not red specs like this before.
dwweiche wrote:
I would assume the number of processors does nothing to improve noise performance. Noise reduction speed would probably be improved.
Also I don't think multiple processors can improve a sensors ISO performance.
You mention there should be improvements over the 50D. Yes. The big one is 8 FPS with an 18 MP sensor. That's a lot of data to move that is only possible with dual processors.
Multiple processors allow more readout channels, which allows a slower readout speed; lower readout speed in turn allows lower electronic read noise. Part of the reason the 1 series has had better read noise performance is its use of dual digics, for double the number of readout channels, and this has now been carried over to the 7D. However there have been other high ISO noise improvements in the 7D -- Canon finally has tackled the issue of banding noise, and the 7D shows it in its high ISO image quality.
The Nikon D3 uses six processors to read out the sensor, which in part accounts for its excellent noise characteristics. The Sony CMOS sensors process the sensor in a massively parallel way, with a readout and digitizer for each column of the sensor; this helps to account for the stellar low ISO noise performance of the D3x, in spite of its high MP count.
Zenon Char wrote:
It seems to be more noticeable with this unit. I just found out about PS Despeckle and applied it for the white specs. Still curious as to what causes it? Here is one after Despeckle. I have not red specs like this before.
It is a sign that you have all NR turned off and the camera is not interfering with your raw files, so they will be really crisp, and respond extremely well to simple noise reductions techniques, like Photoshop dust and scratches: http://cyberphotographer.com/7d/speckless.jpg
mogur2 wrote:
Thanks for all the suggestions for setting focus on the 7D.
Below are a few shots taken with the 7D using the Sigma 50-500 zoom. Cropped to about 1/3 frame.
Not to seem harsh but those look worse then what my G10 does. What iso where those at? The birds look like blotches.
We get that seve lighing here in New Mexico. The contrast is just to harsh to record. I try to always keep the sun to my back when shooting in such light.
The puts the image back into the recordable range.
Zenon Char wrote:
Yes that worked well. What causes that?
Impulsive component to the read noise in RAW data. In other words, a few pixels that are outliers in the noise distribution. They are particular colors (like red) when it is a red pixel in the CFA that has the noise impulse -- the surrounding G and B pixels aren't as noisy, so you get color saturated noise spikes. White is more likely a spiky G pixel, due to the way demosaic works.
