fredmiranda.com
Login

Moderated by: Fred Miranda
  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Canon Forum | Join Upload & Sell

1       2       3              23              25       26       27       end
  

Archive 2009 · 5D2 low ISO poor v 1Ds3

  
 
jonbrach
Offline
• • •
Upload & Sell: On
p.24 #1 · 5D2 low ISO poor v 1Ds3


to add,i just received the prints of the shots i posted and the rest of the set and they are simply stunning....to be able to shoot in the near darkness of a club at 3200 and above and have prints that are so clean and detailed is a revelation...sure we can always find some faults if we blow things up to 200% etc. but in terms of real world use it is wonderful...


Mar 07, 2009 at 03:04 PM
erichard
Offline
• • •
Upload & Sell: Off
p.24 #2 · 5D2 low ISO poor v 1Ds3


Thanks Richard,

Well, I had been putting it off, but after emailing Canon a few weeks back about this matter, they asked about highlight tone protection (HTP), among other things including the firmware, and I told them I'd get back to them on whether it made a difference. So today, I took a few shots, and here is a comparison of the same view, one taken with HTP, and the other with it off. They were processed in Lightroom, and I synced the parameter adjustments via their sync function. So everything is controlled but HTP. The first is with HTP on, and the second without. And a crop of each in the same order follows. Discussion below.

http://gallery.me.com/ord4/100386/_MG_1324/web.jpg
http://gallery.me.com/ord4/100386/_MG_1325/web.jpg
http://gallery.me.com/ord4/100386/_MG_1324-2/web.jpg
http://gallery.me.com/ord4/100386/_MG_1325-2/web.jpg

It's interesting to see what HTP does to the overall picture taken at same f/3.2 1/200th ISO 200 vs the same without HTP. The crops I think demonstrate that the HTP is in fact dramatizing a faint cross hatch, linen like pattern of vertical and horizontal lines in the shadows (I deliberately threw the background out of focus to make it clearer (sounds impossible ;^) I know), and used a bright end of a stick of wood to get HTP in action to protect highlights). The Non HTP image may have a slight hint of the cross hatch pattern, but really, only if you already knew what to look for. My guess is the cross hatch pattern is truly there on both, probably reflecting sensor pixel alignment, but that the HTP magnifies what remains obscured in the non HTP image.

I hope Canon can come up with some firmware to deal with this.

(for those not following, this was done on a 5DII with a 70-200 f/2.8 at ISO 200.)



Mar 07, 2009 at 05:41 PM
LennartW
Offline

Upload & Sell: Off
p.24 #3 · 5D2 low ISO poor v 1Ds3


5DII 70-200/2.8 @ F6.3 ISO200

FullRes: Click

Regarding the AF debate: when shooting basketball, I can feel how it locks on and tracks the subject all the way.



Mar 07, 2009 at 05:54 PM
shirozina
Offline
• • •
Upload & Sell: Off
p.24 #4 · 5D2 low ISO poor v 1Ds3


Like I said HTP looks like it just compresses the highlights and in order to do this it has to push the shadows. You can do teh same thing by underexposing and then adding a 'shoulder' to the curves in the RAW processing software.Turn it off and expose as far to the right as possible. Those 3200 pics look great and everyone agree's that the 5DII had good high ISO capability but this topic is about low ISO image quality.


Mar 07, 2009 at 08:53 PM
thrice
Offline
• • • •
Upload & Sell: Off
p.24 #5 · 5D2 low ISO poor v 1Ds3


HTP is a joke for those who don't know how to meter or use the zone system... or are too lazy to think about the shot.


Mar 08, 2009 at 06:53 AM
Jonathan Wong
Offline
• •
Upload & Sell: Off
p.24 #6 · 5D2 low ISO poor v 1Ds3


thrice wrote:
HTP is a joke for those who don't know how to meter or use the zone system... or are too lazy to think about the shot.


Not sure how it compares to shooting RAW and pushing in post processing. But if you shoot JPEG and/or have strict time constraints on your work such as a sports consignment it would be useful.



Mar 08, 2009 at 07:43 AM
Daan B
Offline
• • • • •
Upload & Sell: Off
p.24 #7 · 5D2 low ISO poor v 1Ds3


thrice wrote:
HTP is a joke for those who don't know how to meter or use the zone system... or are too lazy to think about the shot.


Does HTP affect RAW data?



Mar 08, 2009 at 09:34 AM
erichard
Offline
• • •
Upload & Sell: Off
p.24 #8 · 5D2 low ISO poor v 1Ds3


HTP would be handy in fast changing lighting where you don't have time to review the histogram every time (or don't want to think about it I guess). Sometimes it's just a partly cloudy day with alternating bright sun and shade, at say a soccer game.

I received a response from Canon today (Sunday, believe it or not) saying that with the firmware upgrade, I should not be getting those results in HTP. They want me to send it in, and maybe I will.

As far as Raw data, I'm fairly curious as to what the answer to that is. That picture above of the people walking across the icy lake would be significantly better if I could get rid of that banding. If it's not in the RAW data, that means I could do so, I suppose (but don't know at what cost to the look.)



Mar 08, 2009 at 02:37 PM
FashionBoy
Offline

Upload & Sell: Off
p.24 #9 · 5D2 low ISO poor v 1Ds3


I would suggest sending your MkII in if you are having these problems... it is not normal AFAIK. I personally believe there was either a defective production run or a defective component causing a variety of IQ issues but the newer serial #s don't seem to exhibit any of these problems. I had a very early serial no. that I received in late November that was wrought with issues such as shadow banding at ISO 100.

I sent it in a few weeks ago for service and am waiting for it to come back.



Mar 08, 2009 at 03:05 PM
erichard
Offline
• • •
Upload & Sell: Off
p.24 #10 · 5D2 low ISO poor v 1Ds3


Yeah, I received mine in late January I believe, from Amazon, which could not have had it in stock very long, as it was on backorder off an on for a while. I was rereading the note from Canon, and it is somewhat ambiguous, so I have replied asking clarification. Here's the quote:

"It is not normal to see the pattern you are describing after updating
the firmware in your EOS 5D Mark II. Unfortunately, if the pattern
still appears after updating the firmware and disabling Highlight Tone
Priority, your camera may need to be examined by a service technician.
We apologize for the inconvenience"

Of course my banding doesn't appear after disabling HTP (at least in my brief test), hence the ambiguity (because I did describe it precisely).



Mar 08, 2009 at 03:29 PM
Jonathan Wong
Offline
• •
Upload & Sell: Off
p.24 #11 · 5D2 low ISO poor v 1Ds3


Daan B wrote:
Does HTP affect RAW data?


No effect on the data but the RAW file is tagged as HTP:

http://forums.dpreview.com/forums/read.asp?forum=1019&message=25080916



Mar 08, 2009 at 03:32 PM
erichard
Offline
• • •
Upload & Sell: Off
p.24 #12 · 5D2 low ISO poor v 1Ds3


Thanks for the link. An interesting thread. I think you may be correct, but the only thing that makes me hesitate slightly is there is clearly a difference of opinion on that thread (ie. that ISO 200 HTP ≠ ISO no HTP with PP in camera via tags or whatever). The general upshot of the majority opinion is that HTP could prove useful for jpegs, the prime example being sports photogs who are shooting faced paced sequences, in need of jpegs for size and transmission. The clincher sentence in that thread was that when the OP removed the HTP flag from the EXIF (not yet sure how I would do this), the ISO 100 looked exactly like the ISO 200 HTP. Seems somewhat hard to believe it is that rudimentary, which makes me wonder.

As some have suggested, since most people endeavor to expose to the right, Canon should allow a subliminal preshot (like the 580EX system) in which the sensor detects blown highlights via the curve and adjusts things accordingly. From that baseline, they could allow us to customize what we prefer being adjusted, shutter speed, aperture, ISO, with minimums and maximums, and perhaps linked to focal length, but not necessarily 1:1 with shutter speed (ie. I sometimes like higher than 1:1 for auto ISO). It just seemed this could be automated so much better than it is, given it's simple enough for a chimp to due it (hence the chimping [sic, origin is actually different].)

My guess is that at some point in the future, we will dispense with the mirror system, view the scene via the sensor as in live view, only via a viewfinder of sorts that mimics the SLR viewfinder (blocking out daylight glare, and allowing the camera to be stabilized better with elbows on chest and camera smushed again the face a la the SLR.) That would allow realtime histograms, exposure control, etc. With batteries and LCD's improving (maybe with heat not interfering with more modern LCDs), it may become more feasible. That's what essentially occurs with camcorders, for years now. Come to think of it, the convergence between camcorder and SLR make only be a few iterations away from being complete, the 5DII being a major step in that direction.

The added beauty of that system will be no mirror flop, with its loud noise and vibrations, thereby obviating the need for a rangefinder system, a la Leica. Maybe this is actually where Leica should concentrate rather than on R-10.



Mar 08, 2009 at 04:31 PM
Jonathan Wong
Offline
• •
Upload & Sell: Off
p.24 #13 · 5D2 low ISO poor v 1Ds3


You may induce a lot of noise from constantly active sensor as well as draining battery very fast in live view that way though.


Mar 08, 2009 at 04:59 PM
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.24 #14 · 5D2 low ISO poor v 1Ds3


erichard wrote:
I had the same results. It's not so much the noise, but the regular pattern to it, the vertical banding as if you are shooting through fine linen.

I think it might be the highlight tone priority, but I haven't done a controlled shoot to determine. I wrote to Canon about it. They said to send it in, but I'm not doing it till I replicate the circumstances better.

ISO 200 may be a common ISO, as that's the lowest Highlight tone priority will go.

My guess that's a firmware upgrade away from getting fixed, even though that was supposed to be the
...Show more


i guess im lucky since i have yet to see the vertical banding issue be it ISO6400 or ISO100 or HTP on or off or RAW or sRAW1

your sample looks pretty awful to say he least, that doesnt even look like a deep shadow



Mar 08, 2009 at 05:41 PM
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.24 #15 · 5D2 low ISO poor v 1Ds3


Jonathan Wong wrote:
No effect on the data but the RAW file is tagged as HTP:

http://forums.dpreview.com/forums/read.asp?forum=1019&message=25080916


it does alter raw data in the sense that ISO200 and ISO2oo do not produce the same raw file, but you can get the same affect by underexposing (but you need to completely manualy apply a special tone curve to make it look right then and the jgs look a mess, otoh you can apply say only 1/3 or 2/3 instead of always 1/1 stop htp)



Mar 08, 2009 at 05:45 PM
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.24 #16 · 5D2 low ISO poor v 1Ds3


here is an ISO1600 sample from my copy:

full-image:
http://skibum4.smugmug.com/photos/487552245_ZNHNN-X3.jpg

and then 100% crop with 50% (!) shadow/highlight shadow recovery tool applied:
noisy yes, but this is pulling near blackness and the banding is not that at all vertically that i can see and barely there horizontally (that said ISO6400 under truly dim conditions can show quite a lot of horizontal banding):
http://skibum4.smugmug.com/photos/487551821_cND38-X3.jpg



in case it matters, i'm still on firmware 1.0.6



Mar 08, 2009 at 06:05 PM
brainiac
Offline
[X]
p.24 #17 · 5D2 low ISO poor v 1Ds3


erichards file is looking very familiar to me and skibum's isn't. i want a camera like skibums, not one like mine. i must try going back to 1.0.6 but i am increasingly suspicious that my camera will be going back to canon.


Mar 08, 2009 at 06:48 PM
erichard
Offline
• • •
Upload & Sell: Off
p.24 #18 · 5D2 low ISO poor v 1Ds3


Right. I took some more ISO 100 shots tonight, and the same cross hatch banding is there. So while HTP may magnify the problem, the problem may not be HTP.

I think if people want to test their cameras, ISO 1600 is not the issue. Try ISO 100, in the shadows, pushed a bit for exposure or brightness. The noise is not random, but rather is regular and patterned.

ISO 100, HTP off (not even possible)

http://gallery.me.com/ord4/100386/_MG_1337-2/web.jpg



Mar 08, 2009 at 09:13 PM
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.24 #19 · 5D2 low ISO poor v 1Ds3


erichard wrote:
Right. I took some more ISO 100 shots tonight, and the same cross hatch banding is there, so while HTP may magnify the problem, the problem may not be HTP.



yikes, i guess this clearly puts to rest whether sensors can vary considerably at times
(although perhaps it is more truly called an electronics readout or amp issue than the sensor itself?)

maybe banding, especially vertical banding is like the new lens af calibration before micro-focus adjust....

more and more i am not caring even the hint of one bit that the CF door on mine creaks and squeaks, i'll take that any day over the sensors in the 2 5dmkii above that are vertically banding badly.



Mar 08, 2009 at 09:37 PM
danmitchell
Offline
• • • •
Upload & Sell: Off
p.24 #20 · 5D2 low ISO poor v 1Ds3


A data point, for what its worth, though I don't have any photos to post here yet: Last night I did a bunch of night photography for the first time with my 5DII. I worked at a location that I've shot before - the Mare Island Naval Shipyard in the SF Bay Area - and under similar conditions including an almost full moon and a variety of different types of artificial light.

My basic general approach to night photography is to use the same ISOs and apertures that I would use for daytime photography, and to compensate by extending exposure. Last night I made some exposures at ISO 100, but then figured I'd give ISO 200 a try since the higher ISOs are cleaner on my 5DII than on my 5D.

On exposures of up to 5 minutes the amount of noise in the sky and in shadow areas of the scenes - typically where you'll most easily find noise - is considerably less than what I got from my 5D. I was actually quite surprised at how much better the noise performance was, especially since I was shooting at ISO 200 rather than the more typical ISO 100. I'm usually a skeptic when it comes to the magnitude of the improvements that can be expected from a body upgrade, but I'm no longer a skeptic about this one.

Dan



Mar 08, 2009 at 10:10 PM
1       2       3              23              25       26       27       end




FM Forums | Canon Forum | Join Upload & Sell

1       2       3              23              25       26       27       end
    
 

Welcome back
Log in to your account