Home · Register · Join Upload & Sell

Moderated by: Fred Miranda
Username  

  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Canon Forum | Join Upload & Sell

1
       2       end
  

Archive 2015 · 5DS-R and UWA and HDR

  
 
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.1 #1 · p.1 #1 · 5DS-R and UWA and HDR


Snapsy did a test on the effect of EFCS on subsequent shots and shows the second shot does not get the advantage of reduced camera shake. https://www.fredmiranda.com/forum/topic/1374160

But this is for long lenses. I do HDR brackets mostly with UWA, where subject motion trumps camera shake and where camera shake is not much of a problem on a good tripod. I use a RRS tripod and ball head and a cable release.

But one thing that is good for any focal length is that the first shot gets the EFCS advantage so it makes sense to make the bracket sequence +.0,- The longest shot goes first when EFCS is effective.

In my own test at 11mm I varied from single shots to s burst (silent) to H burst. s Burst is 3 frames per second and h Burst is 5 frames per second. All in live view mode 1. I do not see any shake in either mode at 11mm. I think this works pretty good up to 50mm or so. I have seen shake with my 70-200 when mounted to the lens bracket someplace above 150mm.

Some stuff I learned. If you shoot singles for a blend like HDR, you get terrible subject motion, leaves and clouds do not align and will ghost like crazy. This is solved in mild wind with either the h or s burst mode. Clouds are especially bad.

If you want to get a burst, it is best to AF first and then turn off the AF before shooting, otherwise it tries to focus each time, and UWA lenses do not focus fast. Of course MF would not matter.
I have experimented, and cannot beat live view AF with normal AF or manual focus. If you look at the distance dial on the lens and watch where it ends up after each focus, you will see live view repeats each time, but regular AF or MF will not.

11-24 soft edges:

I have seen very sharp foreground and very sharp infinity with out of focus edges. I started seeing this with my 17TSE back on the 1DS3. But it has become very noticeable with the 11-24 5DS-R. I even saw this with a Ziess 15 f2.8 that I briefly owned when mounted on a 1DS3.

For my 11-24, focusing on an edge about 1/3 to 1/2 up the side at f11 produces best results and seems to solve the soft edge problem I have seen. I use Live view focus with x6 mag, then turn AF off.

This may be a field curvature issue, but it could also simply mean that the left and right edges are further away from the sweet circle of sharpness and need to be more critically focused. The very large DOF for an UWA lens allows sloppy focus to work on the rest of the image.

I also see some cases of fringing on the 11-24 even after Adobes automatic lens correction has been applied. This is in strong light situations. The 17TSE is the better lens in this regard.

I have even seen the soft edge issue with my 24-70 f2.8 mk2 at 50mm. I wonder if the light angle at the extremes entering very tiny photo sites is a problem?

I am finding that ISO50 greatly reduces sky noise for HDR processing. It is still there before NR, but greatly reduced from ISO100 shots. I have heard that ISO50 has less DR than ISO100, but that would not matter for a blended shot.

I have found the 5DS-R files can be pushed pretty far, and I am not seeing the banding I saw in earlier cameras. But I get cleaner results when I blend images that don’t need to be pushed. IE, each area is from an image that
is exposed for that lighting.



Jul 06, 2015 at 09:16 AM
johnctharp
Offline
• • • • •
Upload & Sell: Off
p.1 #2 · p.1 #2 · 5DS-R and UWA and HDR


Canon's BSI tech can't come fast enough, it seems!


Jul 06, 2015 at 10:31 AM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.1 #3 · p.1 #3 · 5DS-R and UWA and HDR


BSI


Jul 06, 2015 at 10:35 AM
johnctharp
Offline
• • • • •
Upload & Sell: Off
p.1 #4 · p.1 #4 · 5DS-R and UWA and HDR


ben egbert wrote:
BSI


Sorry, 'Back-side Illumination', a technology where the circuitry that interfaces with the photo sites on an image sensor is on the back of the silicon substrate instead of in front of it, thus out of the way of the light path to the photo sites.

It's primary advantage is faster readout for stuff like video, electronic shutters, and can improve the rolling shutter effect, but from an IQ standpoint it can also improve peripheral image performance due to the shorter 'walls' around each photo site.

Sony's A7R II is the first 35mm sensor to employ this tech, which they've also used for a while in their camera sensors and 1" sensors, while Samsung is shipping it in their 1.5x APS-C 28MP sensor, and Canon has a patent on the technology, simple as it is, as well.



Jul 06, 2015 at 10:55 AM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.1 #5 · p.1 #5 · 5DS-R and UWA and HDR


Thanks.


Jul 06, 2015 at 10:57 AM
dsjtecserv
Offline
• • •
Upload & Sell: Off
p.1 #6 · p.1 #6 · 5DS-R and UWA and HDR


ben egbert wrote:
But one thing that is good for any focal length is that the first shot gets the EFCS advantage so it makes sense to make the bracket sequence +.0,- The longest shot goes first when EFCS is effective.


I'm not sure that would always be true. The effect of shutter-induced camera vibration depends on the actual shutter speed. If the shutter is open long enough the period of vibrations during the early part is a small part of the total exposure time, and has less effect on the image. Shutter speeds roughly shorter than 1/2 second and less than about 1/100 are most vulnerable. Unfortunately, that's the range that a lot of low-ISO, sweet-light landscape shooting happens in! But in the event that the "long" shot in the sequence will be longer than 1/2 second, it might be best to return to the normal sequence and let the first (mid range) shot get the benefit of EFCS.

Dave



Jul 06, 2015 at 11:19 AM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.1 #7 · p.1 #7 · 5DS-R and UWA and HDR


dsjtecserv wrote:
I'm not sure that would always be true. The effect of shutter-induced camera vibration depends on the actual shutter speed. If the shutter is open long enough the period of vibrations during the early part is a small part of the total exposure time, and has less effect on the image. Shutter speeds roughly shorter than 1/2 second and less than about 1/100 are most vulnerable. Unfortunately, that's the range that a lot of low-ISO, sweet-light landscape shooting happens in! But in the event that the "long" shot in the sequence will be longer than 1/2 second, it might be
...Show more

Going to need to think about this. I have taken a lot of water shots around 1/2 to 1/20 seconds and seen no sign of shake between brackets. But this would have been before I thought about changing to slowest first. The HDR software of course attempts to align shots and does a very good job for most stuff, but clouds and leaves not so well.



Jul 06, 2015 at 12:04 PM
darbo
Offline
• • •
Upload & Sell: Off
p.1 #8 · p.1 #8 · 5DS-R and UWA and HDR


Ben,
Good to hear that you've had success with ISO 50 for cleaner bracketed shots. Naturally it's not a great idea for a single exposure - unless you really need a slower shutter speed with an ND unfriendly lens like the 11-24mm - but, for me, I like my results with ISO 50 for HDR.

I also agree with you that for HDR it is critical to take the exposures as quickly together as possible; that manually depressing the shutter for each bracket could pose more challenges for HDR merges. And, you're right, at 11-24mm the shutter shake may not be a serious concern.

One thought that came to me in all this discussion (here and on Snapsy's thread): I own a Promote Control wired remote (http://www.promote-control.com/), which has more extensive bracketing tools (like automated bracketed shots in bulb mode, automated focus stacking, etc.). The way the Promote Control fires off all the brackets seems different than the way the camera fires them (in 2/10 sec mode), and I wonder if it utilizes the EFC shutter for each and every bracket. I just sent them a support question asking about it and will pass on their answer if it's a good one.



Jul 06, 2015 at 12:52 PM
dsjtecserv
Offline
• • •
Upload & Sell: Off
p.1 #9 · p.1 #9 · 5DS-R and UWA and HDR


ben egbert wrote:
Going to need to think about this. I have taken a lot of water shots around 1/2 to 1/20 seconds and seen no sign of shake between brackets. But this would have been before I thought about changing to slowest first. The HDR software of course attempts to align shots and does a very good job for most stuff, but clouds and leaves not so well.


Well, I'm not saying you are guaranteed to have shutter-induced vibration evident in images shot in that range. I'm just saying that is the most vulnerable range; shutter shockis less likely to be evident outside that range. This is a well-established observation; I didn't make it up!

I'm only pointing out that the concept of giving the longest shutter time the benefit of EFCS may not be the best allocation of resources when that time is beyond a half-second; meanwhile, the "normal" exposure (which is likely to contribute more to the final HDR) may suffer more from being in the vulnerable range without EFCS. And even if both are in that range, I'd tend to give the EFSC to the normal exposure.

You can, of course, initiate the three exposures individually (building in your own two-second delay between each one) while in live view and they will all benefit from EFSC. Its just a pain to do that, and there is more chance of subject motion, I agree.

Dave



Jul 06, 2015 at 02:57 PM
kevindar
Offline
• • • •
Upload & Sell: Off
p.1 #10 · p.1 #10 · 5DS-R and UWA and HDR


Ben, nice summary and points. Only thing that I have to add, the 24-70 2.8 II, is very weak at 50mm in the corners. I have tested this with 3 different copies of the lens, and from f5.6 to f11, both my 24-105, and 24-70 f4IS beat the 24-70 II in the corners.

I dont think this is much due to filed curvature, unless there is additional differential focus shift in the corners, since the lens actually turns in an admirable corner resolution at f 2.8 and f4, Just not tack sharp, and never gets tack sharp.



Jul 06, 2015 at 03:19 PM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.1 #11 · p.1 #11 · 5DS-R and UWA and HDR


dsjtecserv wrote:
Well, I'm not saying you are guaranteed to have shutter-induced vibration evident in images shot in that range. I'm just saying that is the most vulnerable range; shutter shockis less likely to be evident outside that range. This is a well-established observation; I didn't make it up!

I'm only pointing out that the concept of giving the longest shutter time the benefit of EFCS may not be the best allocation of resources when that time is beyond a half-second; meanwhile, the "normal" exposure (which is likely to contribute more to the final HDR) may suffer more from being in the vulnerable
...Show more

I did a test and the subject motion was a killer. I have my camera set up with C1 assigned ISO50 and high speed burst, with C2 at ISO100 and silent burst.The idea is I will use ISO 50 mostly for blue sky scenes where sky noise is an issue and light is probably bright. For all other scenes, I will use ISO100 and silent burst.

I have not actually been on a real shoot with these settings. My prior shoot was with ISO100 and silent burst. But because I was not turning off AF, the burst would often stop to reacquire af. These often show ghosting. But when I get a solid 3 shot burst, I almost never see ghosting.

I have not seen camera induced blur much below 50mm that I am aware of. Not to say it is not there and I am not seeing it. Snapsy's post is making me ready to look for it.

But as long as I am doing HDR, it is pretty clear I need to use burst.





Jul 06, 2015 at 03:23 PM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.1 #12 · p.1 #12 · 5DS-R and UWA and HDR


kevindar wrote:
Ben, nice summary and points. Only thing that I have to add, the 24-70 2.8 II, is very weak at 50mm in the corners. I have tested this with 3 different copies of the lens, and from f5.6 to f11, both my 24-105, and 24-70 f4IS beat the 24-70 II in the corners.

I dont think this is much due to filed curvature, unless there is additional differential focus shift in the corners, since the lens actually turns in an admirable corner resolution at f 2.8 and f4, Just not tack sharp, and never gets tack sharp.


I am forgiving of corners I have seens too many famous prints in galleries with soft corners to worry about it. But I am speaking about edges. Both my 24-70 and 11-24 have produced very good edges when I pay attention to focusing on an edge. Edge focusing does not seem to hurt corners, in my tests center and edge focusing produce about the same corner results. But edges can clean up considerably.

They may still be softer than the center, but the roll off is not so bad.



Jul 06, 2015 at 03:29 PM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.1 #13 · p.1 #13 · 5DS-R and UWA and HDR


Samples of left edge and cloud ghosts

Here are some samples.

1 is an example of a soft left edge with near and far sharp. Focus was at infinity and centered but hyperfocal was nearly the same.

2 is a sample of an in focus left where I focused on the left and stopped down to f11.

3. is the entire image.

4. is a sample of HDR cloud ghosts when shooting 3 single frames with intervals.

5. is a sample of burst mode s shot HDR.





left edge oof.







Left edge as focus target.







full scene







ghost clouds when shooting interval HDR brackets







no ghost when shooting burst mode for HDR



Edited on Jul 06, 2015 at 03:47 PM · View previous versions



Jul 06, 2015 at 03:40 PM
snapsy
Offline
• • • • • •
Upload & Sell: On
p.1 #14 · p.1 #14 · 5DS-R and UWA and HDR


darbo wrote:
Ben,
Good to hear that you've had success with ISO 50 for cleaner bracketed shots. Naturally it's not a great idea for a single exposure - unless you really need a slower shutter speed with an ND unfriendly lens like the 11-24mm - but, for me, I like my results with ISO 50 for HDR.

I also agree with you that for HDR it is critical to take the exposures as quickly together as possible; that manually depressing the shutter for each bracket could pose more challenges for HDR merges. And, you're right, at 11-24mm the shutter shake may not be a
...Show more

I think any wired remote should do the trick - I have a "dumb" non-intervalometer unit on the way. Based on my read of the manual, each press of the wired remote should trigger only a single exposure during bracketing (vs a wireless remote which uses the remote trigger drive-mode that always triggers the full bracket sequence), and since I'd want at least 1/3 second spacing between exposures to settle the shutter from the previous it should work well to manually trigger each exposure. I would do this in LV so each exposure should get the benefit of EFCS. Having a remote with intervalometer support would help automate the process, provided it supports shutter trigger spacing of less than 1 second, otherwise having to wait 1 full second between exposures of a bracket might produce too much subject motion for some of my landscapes.



Jul 06, 2015 at 03:41 PM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.1 #15 · p.1 #15 · 5DS-R and UWA and HDR


I am seeing another thing. At 100% my NEC 26090 shows the image over twice as large as a 20x30 print (100%). For the time being, I need to view at 50% to see what a print might look like. Long term, I need a higher res monitor.


Jul 06, 2015 at 05:04 PM
kevindar
Offline
• • • •
Upload & Sell: Off
p.1 #16 · p.1 #16 · 5DS-R and UWA and HDR


Ben, let me also add that I have many times now resorted to a 2 shot HDR. one ETTR, so so I expose for the highlights, with my focus box there to meter, and and expose until just barely clipping. second exposure is +2.5. My 5d3 allows a 2 exposure shot.
Also if you have not tired the LR 6 HDR tool, I highly highly recommend it.
Its truly amazing, creating a file that has easily over 14 stops of DR, that you can use regular sliders to edit.
The only thing better is blending in layers photoshop, but not all scenes are easily amenable to it. it also has options of decreasing ghosting for you which works pretty well, but limits you to one exposure in that area. so if your shadows have ghosting, it may sometimes choose the shorter exposure, which defeats the idea of HDR.
couple of examples
flickr--5 by kevindar, on Flickr

Cascade 2 by kevindar, on Flickr



Jul 06, 2015 at 05:06 PM
JameelH
Offline
• • • •
Upload & Sell: Off
p.1 #17 · p.1 #17 · 5DS-R and UWA and HDR


ben egbert wrote:
I have not actually been on a real shoot with these settings. My prior shoot was with ISO100 and silent burst. But because I was not turning off AF, the burst would often stop to reacquire af. These often show ghosting. But when I get a solid 3 shot burst, I almost never see ghosting.



Ben, if you use back button focus (I use AF-ON), then you don't have to worry about the camera trying to reacquire AF. Allows you to prefocus using your preferred method (manual, live view ...) and not worry about that focus ever changing.



Jul 06, 2015 at 05:48 PM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.1 #18 · p.1 #18 · 5DS-R and UWA and HDR


kevindar wrote:
Ben, let me also add that I have many times now resorted to a 2 shot HDR. one ETTR, so so I expose for the highlights, with my focus box there to meter, and and expose until just barely clipping. second exposure is +2.5. My 5d3 allows a 2 exposure shot.
Also if you have not tired the LR 6 HDR tool, I highly highly recommend it.
Its truly amazing, creating a file that has easily over 14 stops of DR, that you can use regular sliders to edit.
The only thing better is blending in layers photoshop, but not all
...Show more

Two very nice images. I have used LR and ACR (both work the same ai I dislike the LR UI.It is very good but I use SNS 1.x and its better. Now Nik lest you choose which image to use for primary and handles ghosts very well. But it is worse for sky blotchyness.

My system is working really good. I was not complaining, just passing on some stuff I have seen and my current solutions.

I am a handson guy and don't get real interested in theory, just like when I engineered. When I see something that works, I pass it on. I get conversation and often that leads to better solutions.

I too have seen that 2 shots are often enough. It is so easy to do three and process three or only two, that I probably don't need to worry.



Jul 06, 2015 at 06:05 PM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.1 #19 · p.1 #19 · 5DS-R and UWA and HDR


JameelH wrote:
Ben, if you use back button focus (I use AF-ON), then you don't have to worry about the camera trying to reacquire AF. Allows you to prefocus using your preferred method (manual, live view ...) and not worry about that focus ever changing.


I tried that back maybe the 5D? and never got the hang of it. I read about it recently and thought this would be good. I this a one time thing? IE, focus, push the button and then it will stop refocusing until you are done with the bracket? I don't want to touch the camera after I get focus.

Of course all I need to do is try it and I will for sure. Better that turning off focus on the lens.



Jul 06, 2015 at 06:07 PM
JameelH
Offline
• • • •
Upload & Sell: Off
p.1 #20 · p.1 #20 · 5DS-R and UWA and HDR


The focus only happens when you press the AF-ON (or * if you prefer). Otherwise the camera never tries to focus when you actuate the shutter either with the shutter button or a remote. And not just restricted to bracketing either. You can also leave the lens AF switch on all the time.


Jul 06, 2015 at 06:12 PM
1
       2       end




FM Forums | Canon Forum | Join Upload & Sell

1
       2       end
    
 

You are not logged in. Login or Register

Username       Or Reset password



This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.