March 01, 2015, 03:53:42 AM

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - kaihp

Pages: 1 ... 4 5 [6] 7 8 ... 21
76
I recently discovered on taobao that Benro now also sells filters - specifically also 10-stop ND filters. Since ND1000 filters can help give some interesting effects, I decided to purchase first a 77m and a 82mm filter.

I bought the "SD" (Super Digital) filters. Inspired by Bryan Carnathan's article about 10-stop filters I decided to do a quick test of the filters. As I had previously purchased a 3-stop Tiffen 82m filter (82ND9) I decided to test that also for reference.

Since I don't have an "official" reference color plate, I searched around and downloaded and printed this target: http://photojoes.org/colorchecker.jpg on a laser printer and taped it to a door. Since the printing and my two Osram table lamps used for illumination are likely to give a color cast, so I used the gray boxes to adjust the color temperature for each picture.

(I should say that I'm very much still learning about color temperatures, so I could have done the adjustments incorrectly. If so, please let me know and I'll be happy to update/correct things.)

I chose a deep DOF (f/8) to ensure that I wasn't getting hurt by creases in the paper and non-planarity issues from the taping. Also, since I'm not sure of the lamp filament type, I shot at 0.5sec exposure so reduce or avoid any flicker issues. The 0.5sec, f/8, ISO 100 gave standard exposure with the two lamps.

I'm showing 5 shots, all adjusted to the same (theoretical) in-camera exposure. They have been run through DPP for color temperature adjustment and rescaling to 1200 x 800 px, nothing else.

The camera (5D3) was (obviously) mounted on a tripod (concrete floor), live view was used for initial focus and the lens was then switched to MF. IS was off. Silent shutter and a remote (wireless) trigger was used.

All shots taken at f/8, ISO 100.
1) Control shot, 24-70/2.8L Mk II @ 70mm, 0.5sec.
2) Tiffen ND0.9, 24-70/2.8L Mk II @ 70mm, 4sec.
3) Benro ND10, 24-70/2.8L Mk II @ 70mm, 480sec.
4) Control shot, 70-200/2.8L II @ 200 mm, 0.5sec.
5) Benro ND10, 70-200/2.8L II @ 200 mm, 480sec.

So what are my conclusions so far?
Like Bryan, I find that the Tiffen gives a really bad color cast and much reduced contract. Sharpness seems OK, though. Cost is USD136, according to their website
The Benro filters look quite good with colors very close to the control picture (after correction) and little or no sharpness lost. I'm quite happy with what I see here, especially with the price paid: RMB485-600 (USD88-96) .

Comments, suggestions and improvements/corrections to methodology / design of test are welcome.

77
We just got back a damaged pelican case at work.... Air Canada ran a forklift through it. Fortunately, they hit the box with the tools, not the one with the 60Ghz spectrum analyzer.....

 :o Jeheezuz, forklift ramming a 60GHz spectrum analyzer would have been expensive.

78
I just now received  an update for my 50 Art at the repair shop.

They have ordered the mother board of the lens to replace. I can guarantee, as an old consumer electronics rep guy, they have NO idea what's wrong, and decided to switch out all the electronics and cross their fingers. It's also cheaper to replace the whole board, than to find a single component that might act up, very tricky to find.

I returned mine yesterday. Sigma called me today and said they would just replace it with a new lens. It took them 10 minutes of testing to determine that it was totally sick.

I think these two reports are (relatively) good signs - on Eldar's copy Sigma is able to immediately reproduce the errors and "declare it sick".

In both cases, I seriously doubt that the service shop have the debugging tools to determine why the lenses are exhibiting these errors. For that, they most likely will have to ship them back to the factory or R&D. I seriously hope that they will ship them back (to R&D) for analysis, instead of just scrapping them locally.

79
Btw, when you guys post a working it non working 50 Art, can you please share which body you are using it on, and what sorts of values you have your afma set too?

I have two buddies with 50 Art, one using 1dx and one using a 6d and the one using the 1dx has hm the same issues I and Eldar has, but the one with the 6d it seems to work fine.

First thing I will do when I get it back is to test it in the 5d2 for a while, I hope it's not a 1dx issue.

Good point, Viggo.

If I recall correctly, wickedwombat is shooting with a 5D3 (his lens is OK), and Eldar has used both a 1Dx and a 5D3 (both have problems with his 50A, it seems).


80
1D X Sample Images / Re: Any Thing shot with a 1Dx
« on: May 18, 2014, 02:33:41 AM »
Timbavati Sth Africa

1Dx 400f/2.8 II @ f/2.8 & 1/500th ISO200

Wow, that leopard looks like it's lounging deliberately there for you to take beauty photos of it! :D
Just gorgeous, Edward.

81
And when it misses it still locks on the focus point you use, so you don't know until you review the image on the screen that it missed. It doesn't hunt at all, it locks fast, just in the way wrong spot.

I know nothing about AF algorithms etc, but this seems odd. Wouldn't the camera do several Phase-Detect rounds to assure that focus is achieved? Initial PD-AF to determine focus direction and distance, and then at least a follow-up to confirm that focus is achieved? I recall Roger Cicala writing about this on his blog some time ago (a search reveals this was in July-August 2012).

Yeah, modern-style AF is closed-loop, so the camera should confirm the focus. And yet, it doesn't change away from the incorrect focus setting. This would indicate that the lens thinks it's in the position where it's supposed to be, and yet, it isn't. Wombat is probably right: some physical/electrical part failure seems likely.

Oh, and gratullerer med 200 års dagen til Norge (selvom det var i går) til alle Nordmenn her i tråden :).

82
I trust sigma will sort it out
I'd be interested to find out what specifically was wrong with your copy though

I doubt its software otherwise everyone would have these problems

I'm beginning to suspect internal connections like (flex)cables, connectors etc. This could very easily give erratic behaviour. Hmmm; come to think of it, ESD damage can sometimes also give very odd behaviour.

I would also be interested in knowing the root cause of these problems that Viggo and Eldar has.

83
Hehe, yeah they are easy to approach, but they can take a swing also. My youngest wanted to give it a hug, got within half a meter before it gave her the evil eye and hissed, lol. My daughter yelled at it and I had to pull her away, she wasn't very scared of it.

I take your daughter isn't very old, since she's not scared on a swan? - In Copenhagen City, swans will lie on their eggs in some of the lakes, but they pack a pretty nasty bite - not that I've tried it personally, but from time to time there are stories about teenager boys regretting that they were taunting a swan too much, getting too close for comfort.

A former cat of mine tried to take on just a duck once. All he got was a nasty wound on top of his scull. He never tried a second time - ducks could walk past him within one foot and he would pretend not even to see them.

84
Canon EF Prime Lenses / Re: Canon EF 300mm f/2.8L IS II USM
« on: May 03, 2014, 03:57:53 AM »
In terms of DOF, f/2.8 is why we buy the lens, but sometimes it's not needed. The reason I say that is because the photos you've posted all have rather clean and/or dark backgrounds.

Ah, but that's because I'm deliberately shooting when there is a clean background :) There was a tree and a piano propped up in the background the other day  >:( so I choose to shoot when there was good seperation from the model to the background as possible (no trees growing out of heads etc).

Also, have you read the 1D X AF Guide (now in mobile version as well)?

As the two cameras are very similar, it's well worth a read.

Thanks for the link. I read the Canon DLC pages on the 1Dx/5D3 AF system back when I got the camera 1½ year ago, but you prompted me to sit down and read the 1DX AF guide. About Case 3 it states:

Quote
Effective when you want to continuously photograph targeted athletes one after the other
So here's my "model parade" setting. I'll try that next time (now what mall has the next show?...)

Thanks for your detailed input, I really appreciate it.

85
Third Party Manufacturers / Re: Sigma USB Dock does not work
« on: May 02, 2014, 11:04:37 PM »
Thanks, but I had all the latest versions, so that did not explain it. But sometimes computers work in mysterious ways. I had been through all the standard close the software, retry, reboot the computer etc., with no effect. But after coming home from a trip to the wine store(!)

I'm quite confident that Vinmonopolet would like to quote you for saying that a trip there solved all your problems  ;D

86
Canon EF Prime Lenses / Re: Canon EF 300mm f/2.8L IS II USM
« on: May 01, 2014, 11:00:25 PM »
I also realized that I wasn't waiting for the confirmation light.

I'm not sure if any of these might apply, but hopefully you and others can learn from my many mistakes :)

Thanks, mayguyver. I was indeed hoping to not only learn from the CR members like you, but also help others who might be in a similar situation in the future, by starting this up. Maybe I should move this to another thread (this being the Lens Gallery after all)?

My former cameras were 10D and 50D, so like you, I never bothered with anything but the center point before. And so far with the 5D3, it's been nailing the focus in non-action type shots.

1. Not waiting for AF confirmation
I could very well be guilty this. However, with current settings (after yesterday night), I am not get any keepers from the "model parade" at the end of the show. It's on my TODO list.

2. Focusing on too small an object
Ugh. I learned this on the 50D the hard way. It shouldn't be a problem here, but it's good to keep in mind.

3. Moving the lens too much during AI Servo lock
4. Selecting a non-cross point as the initial AF point
No comments.  :-X

5. Fussing with the AF settings too much
This is where I am now :)

6. Not having my lenses microadjusted
I have Focal Pro and AFMA'ing my lenses is about the first thing I do with them. My setup is less than ideal (only about 7-8 EV light, vs Neuro's recommendation of about 11-2EV), but when comparing with real-life photos it looks good enough.
FWIW, my 300/2.8 required +2 without extenders, and +3/+1 with the 1.4x/2x MkIII extenders.

7. Bumping the diopter adjustment
Been there several times  :-[ It never happened to me on the 10D/50D, but the 5D3.... Good thing is that I rely on the visual focus ack (AF point blink), not that it looks sharp in the VF. I cannot do MF well with the standard focusing screen - I need the old fashioned split prism circle to do MF comfortably, so I ended up buying such a screen for my 50D and still considering it for the 5D3 :)

8. Shooting in the dark
I'm shooting ISO 400, 1/640th, f/2.8, so there should be plenty of light (at least at that venue).
 
9. Having AF tied to the shutter
BBF forever :)

10. Depth of field
Hey, I bought the f/2.8 to shoot at f/2.8, not f/4  >:(  ;D
Seriously, you do have a point about too shallow DoF. But this was my main purpose for the shooting: get acquainted with the lens and the AI Servo system to do those f/2.8 shots. When I understand that, I can always go back to shooting f/4.

After going over almost 5.000 shots from the shows during the last week, I think I can add one more thing:

11. Being too picky about perfect focus - while totally nailing the focus on one of the eyes is ideal, many of the not-quite-nailed shots are probably good enough (at least for web).

As for the AF settings, I kept using the 4-pt expansion, which worked well for me.

I kept using AF Mode 1, but tried pushing the Tracking Sensitivity and Accel/decel tracking from their default level (1) to fastest (2). I believe this was a win, but the jury is still out on this one (I did this in the last part of the show, and the models were goofing around during the Ed Hardy show - see below). However, it is for sure not fast enough for the "model parade" like I wrote above. There, maybe AF Mode 3 (instant lock on whatever comes into the AF point), might be The Right Thingtm.

For the first time ever, I tried and appreciated the separate orientation linked AF points - I used center point for landscape, and a 2nd top row point for portrait shots. When I first set it up and tested the changeover speed, I was a little disappointed that the changeover wasn't instantaneous, but required a split-second. But in real life, it's insignificant and without a shadow of a doubt much quicker than I can move the AF point myself.

Here's two shots of the same model, one where the focus not quite nailed and the other nailed.


87
Damn, I had just convinced myself that I don't need a 50mm lens, but really a 85mm. And then Sigma does this ??? :o

Sigma, please hurry up with that 85A!  :)

88
Canon EF Prime Lenses / Re: Canon EF 300mm f/2.8L IS II USM
« on: May 01, 2014, 01:16:54 AM »
Call me backwards, old fashioned or whatever you like, but I don't really trust the idea of the 61 pt auto point selection.
You're none of those things and I absolutely hate 61-pt for One-Shot AF because it always seems to select the wrong point.

OK, I'm not alone here  ;D

For AI Servo, however, you get to select the initial point, and once you place that on the subject, lock on, and begin tracking, the subject is now tracked throughout the entire AF sensor, not just the zone.

The difference between auto selection in One-Shot and manual pre-selection in AI Servo is the key here.

I will have to work on that some more, because the 61-pt auto seems to go "all over the sensor" for me, even in AI Servo - seemingly regardless of my pre-selected point.

But on to my experiences last night:
As suggested, I did a Clear All Camera Settings and Clear All Custom Functions on the body to get rid of the BIF-optimized settings that I've previously tried and set up the camera to my preferences (BBF, direct selection of AF point on the joystick, AFMA enabled, etc).

When starting to shoot, I found that the AF was doing something really weird on me: it would choose the center AF point for final focus even if I had set the AF point to a different point. After desperately trying to change settings on the fly and understand what was going on, I caved in and gave the body another Clear All Camera Settings. That fixed it. Clearly, I don't understand well enough how all those AF settings work together, and I didn't have the time to sort it out the systematically at the time.

I used AI Servo, High-Speed continuous and shooting short bursts of 2-3 shots at a time. I would vary keeping AF engaged while following the model walking down the runway to stopping/re-acquiring AF with the BBF. Same when they paused at the end of the runway.

AF selection points:
When using both Zone and 61-pt auto-selection, I found that it focused on the "wrong" points, just like mackguyver is describing. I didn't find much difference between the 5 & 9-pt expansion modes, where I had more lenience about following the model with the selected AF point, than when using a single point. No surprises here  :)

AF mode and adjustments:
With the resets, I went back to AF Mode 1, without adjustments. Later I tried to change the Focus/Shot priority from both/both to focus/focus, but since this doesn't seem to be registered in the EXIF, I am unable to distinguish between the two settings. Next time, I should shoot to a new card folder when changing modes.

I didn't spend too much time on chimping during the show, but my impression is that chimping is only good for sorting out the glaring misses: I have to get the pictures up on my computer monitor at 50% or 100% to sort out the shots where the focus is almost there vs the shots where the focus is just nailed.

Looking at several series of shots (each model walking down the runway), it looks like that the tracking speed / snap-on speed is not up to my expectations. This could be the start of a longer investigation, since chimping doesn't seem to be good enough to tell the difference.

89
Canon EF Prime Lenses / Re: Canon EF 300mm f/2.8L IS II USM
« on: April 29, 2014, 08:36:35 PM »
I would go with zone over 61 pt auto
If he was shooting with a wider lens, I would agree, but given that he's using the 300mm, he's going to fill the frame pretty quickly and zone AF might be a bit limiting from a compositional standpoint.

Call me backwards, old fashioned or whatever you like, but I don't really trust the idea of the 61 pt auto point selection. But as I've said, I'll give the 1/5/9/61pt modes the test tonight.

Zone AF should be fine, as I can just put the top third in as zone (I'm shooting exclusively in portrait mode at this venue with the 300mm) and this aligns up with the upper body/face of the models as they walk down the runway.

I'll check in on ideas here before heading out tonight.

90
EOS Bodies / Re: New Sensor Technology Coming From Canon? [CR1]
« on: April 29, 2014, 08:25:41 PM »
I think your cost estimate of $1000 for a FF sensor is way off....

No doubt, thus the mention of the per wafer cost being a WAG (wild-asset guess).  It was a relative example, the absolute numbers didn't matter.

Jon's point about different wafer sizes changes the the relative numbers, of course, but the main point remains - there are possible scenarios where a process change could decrease production costs for FF sensors but not for APS-C sensors, because of the differential effect of random blemishes with FF sensors that are 2.6 times the size of APS-C sensors.

I would guess at a wafer cost is in the range of $2000, given the technology nodes in play (180-500nm)
For the differences in cost structure, see my previous explanation.

Pages: 1 ... 4 5 [6] 7 8 ... 21