I had previously mentioned my focus seemed off on mine and had set it to +7. After further review, it was just a couple lenses. Naturally the first couple I grabbed to play around with the camera were the iffy ones. Spent an hour or so going through all my lenses and tweaking them slightly. IMO, the microadjustment is one of the best features Canon has come out with in the last couple years. Then again, I never noticed them being off with my other cameras, so maybe "close" is close enough most of the time. I didn't test the other bodies aside from 1 lens on the 40D. It could use a +2 adjustment, but that's not much at all. I did notice during testing that the Sigma 120-300 was noticeably less consistent focusing than the Canon lenses I have. No surprise there, I can see that in daily use as well.
I had a little visitor during the test. Shot at ISO 100 large JPG from about 6-8 ft away with a 300 F4 IS. 18MP seems to give some crop lattitude. While I'm not ready to hang this on my wall at 16X20, I think it's a good example of the capabilities available.
canon pants wrote:
what is the point of it working only in live mode? wouldn't you want it more?
Canon canada said they have no bulletines on it or anything and that would explain why.
I think that it probably only works in live mode since that's the only time that the mirror is up. My guess is that mirror slap constitutes the bulk of the shutter noise.
ExxWhy wrote:
I had a little visitor during the test. Shot at ISO 100 large JPG from about 6-8 ft away with a 300 F4 IS. 18MP seems to give some crop lattitude. While I'm not ready to hang this on my wall at 16X20, I think it's a good example of the capabilities available.
alundeb wrote:
The banding pattern is consistent between shots from the same camera, it goes away by black frame subtraction, so it is fixable.
It appears to be an issue of calibration of the readout channels. It goes away at high ISO because it's a post-amplification problem -- at high ISO the signal is amplified more and more relative to the pattern noise, so the latter is less intrusive.
It very likely could be taken care of by black frame subtraction, but which popular RAW converters provide such a feature? None that I know.
ejmartin wrote:
It appears to be an issue of calibration of the readout channels. It goes away at high ISO because it's a post-amplification problem -- at high ISO the signal is amplified more and more relative to the pattern noise, so the latter is less intrusive.
It very likely could be taken care of by black frame subtraction, but which popular RAW converters provide such a feature? None that I know.
Emil - Have you seen enough raw files from a variety of cameras to have a take on how prevalent it appears to be? Is it seen in all cameras, and if so, how much variation between cameras have you noticed?
(In practical terms, when I get my camera, push the exposure, and look for it, how do I tell if I have a "normal" or an excessive amount?)
Tom_W wrote:
I think that it probably only works in live mode since that's the only time that the mirror is up. My guess is that mirror slap constitutes the bulk of the shutter noise.
I often use liveview with my 5D2 for precisely this reason. Trigger liveview outside the church, tip-toe in, get a few shots keeping liveview running, and tiptoe out to switch off liveview. It works a treat - just as well, if not better than, the silent drive mode on the mk3's, although they can do this trick too.
ejmartin wrote:
It appears to be an issue of calibration of the readout channels. It goes away at high ISO because it's a post-amplification problem -- at high ISO the signal is amplified more and more relative to the pattern noise, so the latter is less intrusive.
It very likely could be taken care of by black frame subtraction, but which popular RAW converters provide such a feature? None that I know.
Exactly. It is not conveniently handled by now. I guess my comment was meant just to say it is not impossible to fix. It could possibly be handled in firmware as well.
alundeb wrote:
Exactly. It is not conveniently handled by now. I guess my comment was meant just to say it is not impossible to fix. It could possibly be handled in firmware as well.
If Canon can be bothered to do dust subtraction (does anyone use that?), they really ought to be able to address this.
jamato8 wrote:
The quiet mode only works in live mode. I just called Canon.
So, user error, or misunderstanding but it seemed like the quiet mode was for regular shooting but if you look closely in the manual the information is under the "live mode" section.
Well it is still quieter than either my 5D or 5D II.
Makes sence as the mirror has already been flopped out of the way........>
canon pants wrote:
what is the point of it working only in live mode? wouldn't you want it more?
Canon canada said they have no bulletines on it or anything and that would explain why.
The silent drive mode of the 1-series Mark III is what you're thinking of--you can delay the mirror return and shutter recock to spread out the noise instead of it all happening in quick succession. This is one feature I hoped to see on the 7D and was disappointed to see not included.
FWIW, the only time I really see the low ISO banding/pattern noise on my 5D2s is when I have shot with Highlight Tone Priority enabled. With HTP enabled, pattern noise can often be readily seen in the shadows or in corners which have been corrected for vignetting even at ISO 200, in my experience.
Perhaps this is something that is affecting the 7D output too? (I pretty much never use HTP due to this issue. Since I shoot RAW 100% of the time, I can always lengthen the highlight tone curve in post.)
But if some 7D users have HTP enabled, it could be amplifying whatever pattern noise is there (if any).
Yohan Pamudji wrote:
The silent drive mode of the 1-series Mark III is what you're thinking of--you can delay the mirror return and shutter recock to spread out the noise instead of it all happening in quick succession. This is one feature I hoped to see on the 7D and was disappointed to see not included.
That is what I thought it would be. oh well. not why I bought it.
I went out and shot a few birds with the 200 f2 2X TC II. The fps is really nice and the focus was very good. I haven't processed the images yet, just getting a handle on the flow of things but the fps is really appreciated and the shutter sound is also much improved.
jamato8 wrote:
and the shutter sound is also much improved.
Canon knew this aspect of the 7D was good enough to be advertised.
My sister-in-law, who bought my 20D years ago, might now buy my 50D. I was mentioning the quieter shutter to her. 'I kind of like the substantial shutter sound of the 20D,' was her response. Well... I never did.
paulfeng wrote:
Emil - Have you seen enough raw files from a variety of cameras to have a take on how prevalent it appears to be? Is it seen in all cameras, and if so, how much variation between cameras have you noticed?
(In practical terms, when I get my camera, push the exposure, and look for it, how do I tell if I have a "normal" or an excessive amount?)
I've not looked at a huge number of RAW files, but GaborSch over at DPR has and it seems to be present on most of them. There does seem to be sample variation as to the strength of the pattern, as one would expect if it were a calibration issue.
Gabor has posted a number of examples using his RAW tool RAWAnalyze, you could take a low ISO blackframe and put it into that tool with the same settings as the samples posted at DPR and see whether your camera is better or worse.
My sense is that this problem is not as nasty as the banding on the 5D2; at least, if Canon were motivated they could probably fix it in firmware, since the pattern is fixed and regular rather than the random banding noise that afflicted earlier cameras, which cannot be fixed in firmware since it changes from shot to shot.
godfather wrote:
I just hold the plate up to the new camera and start removing / reducing / elongating metal until the bracket fits and will not rub on the camera. This L plate (for the 20d) took 15 minutes as I was leaving on a vacation...I still need to smooth it out and add some paint.
FYI FWIW: I just posted here about a couple of (insignificant) dust bunnies on a brand new 7D that just arrived.
People have complained in the past about their camera's showing up dirty, but this looks pretty darn decent to me. People getting their camera's may want to shoot a similar series as I did if you are interested in a baseline.
thw2 wrote:
Just tested my copy using Rawnalyze. No low ISO banding issues.
Seems that it's kind of isolated to a few copies.
just tried pulling deep ISO shadows, wow surprised.
i thought all the pre-release talk had been they had gotten a handle on banding.
looks far, far more obnoxious than doing same to a 5D2 shot (which had gotten lots of flack for low iso banding)
hmm your copy is free of it though?
i looked at the DPR thread and one guy did a quick thing with IRIS and thresholds and mine looked close his sample #1 and a fair deal worse than his sample #2.
is the word that most are good and only a few are like sample #1?
While I'm not ready to hang this on my wall at 16X20, I think it's a good example of the capabilities available.
