-
Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.
NeoGAF Camera Equipment Thread | MK II
- Thread starter MRORANGE
- Start date
No one really knows when. But just wait.
JadedWriter
Member
Is it just me or do not all lenses handle iso noise the same way? I think my Sigma lens handles higher iso better than my Nikon 50, but it might just be my mind toying with me.
Astral/H3X
Member
If it's allowing it to pull in more light then certainly.
What I've learned, studying more about astrophotography, is that if a higher ISO allows more light to get in, then it will actually end up being *less* noisy in a lot of cases. If an ISO800 scene is dark, and you adjusted nothing except for the ISO to say, 3200, you'd actually have less noise in the 3200 ISO image, in a lot of cases. There would be noise introduced by being at 3200, *but* that extra sensitivity apparently cuts out other forms of noise. It's crazy.
EDIT: You may also just be using the two lenses a bit differently without realizing it. If you took a few shots using the Nikon lens and said "wow that's noisy I shouldn't do that ever", but then ended up taking a shot with the sigma at high iso in a better lit environment, then you'd see a difference.
JadedWriter
Member
I guess. I've never really used my 50 at 3.2 at 640 iso but the Sigma 2.8 17-50 seems to be handling it pretty well actually. Only thing I can see being the difference is the optical stabilization on the Sigma, but I've gotten some really good shots with that thing last night at 3.2 Iso 1250 and 1600 that I wouldn't have tried with the 50mm as well.
I was using the 20mm 1.8 on the nikon @ f4 and the 17-35 f/4 on the sony @ 20mm f4 both at 1/50sec. 24fps 1080p and 4k for the sony.
This was a very dark scene on purpose to test the limits, nikon was set to flat profile I and believe slog2 on the sony.
I shot some more "normal" low light scenarios @6400 on both and the sony still spanks the nikon.
Is that also taken iso-invariance into account for the said sensor/body? Assuming you are going to be boosting iso800 image to match iso 3200 image for comparison.
Astral/H3X
Member
I'm still looking into all the factors, this was actually something I found out the other day, but I took a test shot with the same overall exposure at two different ISOs. I took one at 800ISO, 1/30, and one at 3200, 1/125 (the closest I could get to a perfect match), and the noise levels were VERY similar. I turned off all noise correction in Lighteoom. At normal viewing levels I really couldn't tell the difference.
There WAS more noise on the 3200, but not nearly what I was expecting. It seems to lead into the idea that underexposing an image at a lower ISO won't give you a less noisy image.
Granted, a proper exposure at a lower ISO is preferable, but if it's between an underexposure at a lower iso vs a good exposure at a higher iso, it seems the higher ISO actually wins.
It's weird and I'm still trying to perfectly understand it, so I may not be spot on, but that seems to be the gist of it.
andythinkpad
Member
New X100 comes out every 1.5 year. Obviously, the next one will have the 24mp sensor.
TerryLee81
Member
And a new lens, hopefully.
With an AF joystick and a physical ISO dial too.
John Kowalski
Banned
Pff, excuses. OG still serving me like a hero.
look up "shooting to the right" or "filling up the right side of the histogram"
Astral/H3X
Member
I've *also* looked at this, and have found conflicting opinions on whether, to that level, it's worth it. It's very easy to blow out highlights with that, for not much gain after a point.
To clarify, I've found that the difference in noise going from properly exposed, to nearly overexposed, is very little, whereas you're quite likely to overexpose and lose detail.
It highly depends on the sensor's characteristics and how the manufacture messes with the resulting signal (even if shot in raw). To some degree, it also depends on the ISO value the camera tells you. (i.e. is it really 100 when it says it's 100?)
http://www.dpreview.com/articles/7450523388/sony-alpha-7r-ii-real-world-iso-invariance-study
Can't say I understand how it works down to the metal level but it's kind of useful information to know for the camera you use. This guy did a lot of signal analysis (and I mean hard-core stuff if you go to his blog) on A7RII, so here's an example of it.
http://www.dpreview.com/forums/post/56561672
When you underexpose you give up dynamic range. When you don't capture enough light your "image" is hidden in the noise floor so attempts to raise the exposure in PP just amplify the noise along with the image.
Should also note that video is a whole different beast than photos.
The D810 is taking its 36mp sensor down to about 2mp for 1080p. In theory this should lead to less noise, but something in the scaling algorithm is wacky. The Sony is scaling 12mp to about 4mp, and is doing an amazing job.
If I did still photos I would expect the D810 to be much closer to the A7s ii but I think the sony would still win.
Should also note that video is a whole different beast than photos.
The D810 is taking its 36mp sensor down to about 2mp for 1080p. In theory this should lead to less noise, but something in the scaling algorithm is wacky. The Sony is scaling 12mp to about 4mp, and is doing an amazing job.
If I did still photos I would expect the D810 to be much closer to the A7s ii but I think the sony would still win.
The A7s and A7s II sample the full sensor for video, so they're getting the full light-gathering effect of having a full-frame sensor. All other full-frame cameras skip lines while recording video, and fewer lines sampled means more noise. The proportion of lines actually sampled vary from one camera to the next, but the worst-case scenario would be to only sample 1080 lines, meaning on a 36mp camera like the D810 you may effectively only be taking in less than a quarter of the light hitting the sensor, and noise would be expected to increase by two stops because of that.
Yeah that is my suspicion. IDK if its straight throwing out lines (it does pretty well on aliasing and moire that would be apparent in simple line skipping) but its far worse than if you took a 36mp still at the same ISO and downsized it to 1080p.
Astral/H3X
Member
I whipped out my camera, and did a few tests, and it seems that the a6000 also is relatively ISO less. I did notice some differences between ISO100-ISO800, but ISO800-6400 looked identical across the files in noise.
I'll have to check more thoroughly, it seemed that ISO100 was different from 200 etc. But at least now I know that 800-6400 is identical.
EDIT: I do know that one of the "super awesome features" of the A7SII is that it is 4K without "line binning", which sounds like what you're talking about where it only reads every so many lines. Thinking about that now, that would perfectly explain the massive difference between the two.
JadedWriter
Member
Is there anything people do here to test out a lens. Ever since getting this sigma I've been taking it to the night time ice skating rink. It's like the perfect test cause it poses a challenge to everything. It tests out the low light capability, iso handling and auto focus speed.
Rentahamster
Rodent Whores
Depends what I'm using it for. If I'm testing for focus accuracy, I'll use a ruler or a focus testing chart. If I'm using it for low light stuff, I'll go make some dinner plans with friends at a dark restaurant and use it there. If I'm testing for sharpness, I'll do some landscape shots or still life shots on a tripod at several aperture ranges and then pixel peep.
JadedWriter
Member
Well those make sense. I'm not a photo mathematician so I just use it in a way that I enjoy and just examine shit afterwards.
Astral/H3X
Member
I would get a vintage macro lens. When you want to use a macro lens, I've found, you'll *know* you want a macro lens. It's really not something I just keep with me, for example, I go out with the intent of using it.
And, to boot, AF is basically useless for macro. Utterly useless. So no reason not to spend $80 and get a vintage macro lens.
Rentahamster
Rodent Whores
That's the main point, yeah.
JadedWriter
Member
Ok. Sounds like everybody has their own different methods. Works for me.
Depends on why you think you need macro lens I guess. If pseudo macro is enough, which that lens does a pretty good job, then there's no need. You will just get slightly better magnification and flatter field.
Otoh, if you need longer working distance then that's a different story.
Joel Was Right
Member
I tend to think ergonomics are more important to the shooting experience than the level of noise beyond ISO 800. For example, the X series cameras (except the XT1) are not comfortable to carry with the hand due to the lack of a protruding grip. The x100 is the biggest offender of this.
Whilst good image quality is assured in all cameras, handling is not.
Whilst good image quality is assured in all cameras, handling is not.
Re: the X-T1, the handheld ergos are *vastly* improved by adding a thumb grip, though the pricing on the go-to choice is slightly outrageous. But kind of used to that for specialized aftermarket photo/video gear. Feels totally essential after installing though. Other options on eBay are much more reasonably priced and there might be a comparison article somewhere that evaluates them, but can vouch for the Lensmate if $60 tiny bits of plastic and silicone are your jam.
TerryLee81
Member
Do you need the 30x zoom?
If not, I would go for something like a used Canon s110 or s120, with a brighter lens and a bit bigger sensor.
Best would be to look for a used rx100 I, but they are definitely more expensive.
Astral/H3X
Member
I'm starting to realize that I don't particularly like the "normal" focal length. I guess it just doesn't mesh with my style well. Every photo I take with it just feels... Plain. There's some exceptions here and there but overall I usually find myself wanting wider or longer and rarely being happy with normal.
I have the X-e2, and I don't really have any issues with it, I have seen a lot of people add the thumb grips, but I've never had an issue
My issue is that fuji doesn't really have a fast pancake lens for the XT1 to compete in size with the x100. 2.8 is too slow, yes. (My main lens is a 35/1.4 on my 5D2.) So the X100 still has my eye for something that's both compact and fast.
TerryLee81
Member
Damn, the IQ of the 300mm seems to be superb.
The owl picture is handheld at 1/10sec!
http://robinwong.blogspot.my/2016/01/olympus-mzuiko-300mm-f4-pro-lens-review.html
The owl picture is handheld at 1/10sec!
http://robinwong.blogspot.my/2016/01/olympus-mzuiko-300mm-f4-pro-lens-review.html
Astral/H3X
Member
Just went through this on my phone, and even without looking at the crops I'm incredibly impressed.
Rentahamster
Rodent Whores
That....is impressive.
It does look good.The image stabilisation is ridiculous. I know the guy works for Olympus but he could be a little more accurate in his reviews, some of the stuff he says is factually incorrect. The Nikon 300mm F4 weighs less, has the same minimum focusing distance, and costs less money. The Canon version is the same with only the minimum focusing distance being slightly longer. He would be better off just talking about why it is great than trying to compare it to the competition.
jongkookie
Member
I hate myself.
I saw a Nikon 28mm f/3.5 on eBay for $40... that's probably the fastest I've been in trying to hit that "Buy It Now" button lol.
Now I've got primes in 28, 50, 85... I just really need a full frame body then I think I'll really be done with GAS(gear acquisition syndrome) lol.
I saw a Nikon 28mm f/3.5 on eBay for $40... that's probably the fastest I've been in trying to hit that "Buy It Now" button lol.
Now I've got primes in 28, 50, 85... I just really need a full frame body then I think I'll really be done with GAS(gear acquisition syndrome) lol.
Art Teitlebaum
Member
Hhhnnnnnnnngggg. This is really spectacular. And I am tempted to get one just for the fun of doing really sharp tele pics, but I'm simply not professional enough to justify the price tag and my favourite scenario always was street and indoor photography and not wandering in the freezing cold to shoot some birds I don't know anything about. ;-)
I do photo as a hobby among several others and 2,600 EUR is a little bit OTP for a bloody hobbyist as myself.
(But the IQ surely is absolutely breathtaking for that system, goddamnit...)
JadedWriter
Member
I'm right there with you when it comes to GAS, I'm going to need to start buying lights soon. I wonder what's a good affordable starter kit. Don't even get me started on Full Frame, damn I want one so bad just cause of improved iso handling.
You guys are probably more hardcore than I am about cameras but I need assistance.
I would like a camera recommendation.
Not looking to go broke on a camera.
I'd like the best image quality one can get for:
-Around 300 USD (willing to do 400 but no more).
-Smaller in size; I want something that I can carry everywhere all the time.
I would like a camera recommendation.
Not looking to go broke on a camera.
I'd like the best image quality one can get for:
-Around 300 USD (willing to do 400 but no more).
-Smaller in size; I want something that I can carry everywhere all the time.
robin now works for Olympus?
I remember when he was just a fan with a blog.
The lens looks spectacular, would expect nothing less from Oly glass, they make some of the best around. Its still too expensive though.
It would be nice if he did some longer tests. As a user of longer glass, i can say being that close to an owl in real life is almost never going to happen for most people.
Art Teitlebaum
Member
Canon S110, Fuji X20, Lumix LX5 come to my mind. Think about buying used or refurbished then there might be some more choice.
Edit: Olympus XZ-1 / XZ-10 could also be around 300.
Astral/H3X
Member
It'd be pushing up to the $400 mark, but I've heard surprisingly good things about the RX100 line, and the older ones can get down there. Would probably have to be used but definitely something to look for.