Camera manufacturers are always pushing for high iso performance, but why don't camera sensors usually go lower than 100 or in some cases, 200? the lowest base iso you can get on a digital camera is 64 on the d810.
Why don't we have any digital cameras that can go to say, iso 25?
>>2842293
Because why would you need that?
>>2842293
Because it's cheaper to use ND
>>2842294
Dynamic range and color bit depth tend to go hand in hand with lower iso.
>>2842300
No, they don't. They tend to get worse with signal amplification. They don't get better with lower ISO.
The minimum iso available gives you the unamplified signal from the sensor, going lower that that wouldn't make any sense.
>>2842302
Uh, yes they do.
Dmf cameras have 50 and lower ISO available.
When you see a base ISO of 200, it's not because they've given more amplification to the minimum setting, is because they designed a sensor that gathers light more efficiently. If they could affordably lower isos, they would. But these people are running a business to make a profit, so not everyone is willing to use rare custom sensors when they can just pick a sensor out of an industry catalog.
>>2842309
Oh, yeah good point, I thought I knew what I was talking about, but you've shown me the light.
>>2842311
But you don't know what you're talking about.
I bet you don't even know what dynamic range is.
>>2842294
Lower noise, longer exposures, etc.
>>2842307
I'm not talking about artificially lowering it. that's what they do now when you use iso 50 on a digital camera.
>>2842302
but you're not amplifying the signal more if you shoot at a low base iso... Extended dynamic range is one of the reasons why the d810 has iso 64 as it's base instead of 100.
>>2842310
so I guess what I'm wondering is, what exactly determines the base iso of a sensor? Why is it simpler to make a sensor that's natively sensitive at 100 or 200 instead of something lower? I don't understand the technology of what gives it that sensitivity.
>>2842313
Wow, I haven't tried to have a conversation with a 13 year old who thought he was right in a long time. I haven't missed it...
Best dynamic range and lowest noise in an image come from being at the base ISO, because it isn't artificially amplified with electricity. Using a lower base ISO won't do anything, because it will still be the base ISO. The degradation comes from the amplification used to get to higher ISOs. So the same degradation that comes from changing from ISO 100 to ISO 400, would affect the change from ISO 25 to ISO 100 in the exact same way.
It's like you're suggesting that you hold a measuring stick vertically in front of you, and say that you can make the stick longer by holding it up higher, because look, the top of the stick is higher than it was before.
this fucking thread
64 is a native ISO, works fine or sometimes better than 100. there is no 50 native ISO's and they are all completely useless unless for a longer exposure, but 64 on the nikon's can actually give you better lighting a lot of the time
>>2842293
Base/native ISO is when the signal from the sensor is not amplified. Consider two cases: using a (base) ISO of 100 and 50. The camera with a base ISO of 50 needs to gather twice as much (1 stop) worth of light to make the same exposure as the ISO100 camera.
If you aren't familiar with the "photon well" analogy of digital sensors imagine each pixel as a bucket that is collecting rain water (photons). If say a base ISO of 100 made a perfect exposure, metering for an ISO of 50 would require twice as much light and thus more photons would need to be gathered. However, these buckets collecting photons can only hold so many, so if you're dumping twice as much light into them some will inevitably get full (overexposed and bloom).
You can imagine cameras with lower base ISOs as having the ability to gather more light before their buckets overflow. Remember that ISO doesn't affect how much light is actually let into the camera, but it does determine how much is needed to make an exposure.
The d810 with a base ISO of 64 is often said to have a higher dynamic and color range than cameras with a base ISO of 100. If you want to make the same exposure at ISO100 and ISO64, you will need more light on the ISO64 sensor. I think the general idea is that more photons = more data to work with, and more light received from the shadows.
>>2842327
The voltage fed to the sensor determines the brightness. A higher voltage, and you get a brighter image. Too bright, and it exceeds the brightness capability of the Analog to Digital converter, thus highlight clipping. The sensor however, has a minimum operational voltage, under which it won't function properly. The brightness at this voltage (or usually a bit above for safety) is your minimum ISO.
Does that help?
>>2842330
>there is no 50 native ISO's
this fucking post
do your research guy, phase one has native 50, so does Alpa/PO, Mamiya/Leaf, and others.
>>2842329
This is basically correct.
It's also worth pointing out that there's not much need for super low ISOs because shutters are so fast these days. A lot of why ISOs like 25 exist for film is to allow daylight shooting with cameras that have max speeds of 1/500 (i.e. Hasselblad V, Mamiya TLR) or slower.
That's not a big issue now that DSLR/MILC shutters are doing 1/8000, and will matter even less as electronic shutters work their way into the camera world and allow basically infinitely fast shutter speeds. In another few years we'll probably be able to shoot straight into a welding arc at f/1.4 and still expose properly.
>>2842329
Except this isn't true, if you meter for an ISO of 25 you will need two more stops of light than if you metered for an ISO of 100. Suggesting that there is no difference when a camera has 4x as much light to work with is silly
>>2842337
There are two ways to lower ISO. Either lower the voltage that the sensor works on, or increase the amount of light that it can gather. Obviously the second is more desirable, but it's way more difficult and expensive, thus the more realistic option is to lower the voltage. That doesn't mean more light, just less power, which does hurt your higher ISOs, even 100 in this theoretical instance.
>>2842337
So still, you don't understand how things work, and you're just using logic and photo 101 knowledge to try to suggest that you're right.
It's not the amount of light that's the issue. It's the way the camera handles the light. It's the DAC (which is the current bottleneck in terms of dynamic range) and the way that higher ISOs are handled (More electricity is cranked through the sensor, which causes noise). It's not the lack of light.
>>2842339
>you're just using logic
Oops.
See below.
>>2842338
>or increase the amount of light that it can gather... but it's way more difficult and expensive
Oh so here's an answer to OP's question.
There are people talking about two different things in this thread. As pointed out by >>2842300 and >>2842331, dynamic range and color depth might be associated to a lower base ISO IF we're talking about lower base ISO using higher light gathering capabilities.
>>2842342
Except you can increase the light gathering capabilities without dropping the base ISO from 100. This happened when we jumped from 11 stop dynamic range to 14 stops. There is nothing inherently "better dynamic range and color depth" about a lower base ISO.
A CCD chip is basically an array of miniature capacitors. They only have a limited capacity, which limits how much light they can be exposed to before saturating, and also when a bin is saturated it has a tendency to leak charge to surrounding bins, causing blooming.
>>2842342
correct: the post
Unfortunately companies don't simply user more expensive, challenging sensors because it's not the consumer which will pay more, it's the company producing the cameras. The final price of the camera is based on the market, not the materials used. This is why Canon cameras cost as much as other brands despite their old, poorly performing cheap parts.
>>2842346
That explains why all current cameras have a base iso of 400 these days, since it's all the same lmao
>>2842372
What?
>>2842373
Sarcasm must be lost on you, along with a basic understanding of electronics.
>>2842380
No, I'm just not sure where you pulled "base ISO of 400" from. Base ISO on digital cameras has almost always been very near to 100, and has stayed there, through many years and many improvements to both dynamic range and color.
When Sony gave us 14-stop DR cameras, up from 11, they didn't do it by dropping down from the standard ISO of 100. The base ISO stayed 100 (or close to it) because a lower base ISO doesn't have anything to do with dynamic range or color.
The base is 100 (or sometimes 160, or 200, and sometimes 64, or 50) because that's where people want it to be. ISO 100 was pretty standard for film, so they carried it over. It's a good speed for daylight stuff, and much lower wouldn't be very useful for anyone not doing long exposures, in which case something less like 50 or 25 wouldn't make much of a difference, and they'd rather just use a 10 stop ND filter.
>>2842387
>Base ISO on digital cameras has almost always been very near to 100, and has stayed there, through many years and many improvements to both dynamic range and color.
Canonfag spotted. ISO 200 has more typically been the base ISO for Nikon, Pentax, Sony and Fuji bodies.
Only recently, within the last generation or two, have they started dipping their toes into the slow, cool, deep waters of lower base ISO.
Forget all this 64 base ISO discussion, what's the camera with the highest base ISO?
I've heard rumors that the a6300 has a base ISO of 800 (as in, anything below 800 uses ISO 800 sensitivity and scales the data in software).
EE here
Please kill yourselves
Base/native ISO is the ISO of a sensor at which the analog signal is not amplified before the conversion to digital.
Imagine your camera has a base of ISO 400.
shooting at ISO 100 is just effectively 'turning down' this ISO 400 output. You shouldn't see a marked reduction in noise, only luminace, which needs to be made up for with a larger f/stop or longer shutter speed.
shooting at ISO 800 means you need to amplify the ISO 400 output, which in turn amplifies any errors in the signal giving you noise.
As sensor technology gets better, base ISO's should rise, not fall.
>>2842555
>I have the answer but won't share cuz i'm speshul
Thanks for your contribution, faggot.
>>2842568
>As sensor technology gets better, base ISO's should rise, not fall.
Except the exact opposite has proven true.
While ISO does have something to do with ADC gain, ISO is generally a meaningless number because manufacturers try to make it so that ISO "100" will provide similar exposure across all cameras and film given the same shutter speed / aperture.
So ISO 100 on one camera and ISO 100 on a completely different camera will have different levels of gain in order to give the same exposure.
That's why ISO 50 on any phone camera still looks like actual shit, pixels are tiny so gain needs to be higher to get an acceptable exposure given a 1/30 sec shutter speed, or whatever.
You can't really have discussions about "ISO" or talk about 'decreasing it' for this reason.
Canon 6D goes down to iso 50
I rarely use it except when I forgot my ND filter at home and want to get wide open in sunlight. Still isn't enough sometimes
>>2842680
that isnt fun
nobody says I S O
>>2842680
Tony pls go
>>2842705
>>2842692
>>2842680
Damn dude these are super nice
Especially that one of that building. Do you have any suggestions as to how to practice this? You're getting such a uniform amount of light with barely any flares and nice colors. Is it pretty much all in the placement of where you are so there's no direct light shining at you combined with a lot of PP to get the colors right?
>>2842705
https://youtu.be/OQ1-lTg6AoQ?t=28m4s
Every
fucking
video
>>2842680
wait where in orlando is this
lived there for two years and all I got out of it is that no two people talking about orlando are ever talking about the same place because it's too fuckin big
>>2842335
I think the 5dm3 has a 50iso option too?
>>2842293
From my think (with no research) its probably cause if the sensor has base ISO of say 25, the signal amplified to ISO 100 would be worse than a sensor that's base was already at 100.
Of course this is just speculation. I sure you can research it and find out.
>>2842643
Yeah when I'm using my 100-400mm at race tracks I really wish I could drop to 50ISO.
INB4
>use a filter
Filters absolutely fuck the iq on the 100-400
>faster shutter speed
I want the wheels to not be frozen, so have to stay below 1/500th
>close down aperture
Lots of race track shots look way better with a shallower DoF
>>2842680
ISO stands for International Standards Organization, and can be pronounced any way you want. It's like NASA or FBI.
>>2842787
It's not native, it's pulled in post. It's shot at 100.
>>2842834
Wat?
How the fuck does anything in post change the exif data of a photo you mong?
Cropping a photo doesn't change the focal length.
>>2842789
>Filters absolutely fuck the iq on the 100-400
Then get a better filter... Even A moderate quality one won't be noticeably detrimental. But if you're too much into being autism for it, you should absolutely be able to pull your exposure down two stops in post on anything but a Canon sensor.
>>2842839
Yes it will.
Go read any discussion about the original 100-400mm. Even top of the line filters mess up the iq.
The only positive I can take from your reply is that it just confirms to me the amount of absolute bullshit being said on here.
>>2842842
You don't want to mention at least WHICH 100-400? You said you wished you could drop to ISO 50, and most canon cameras can do that, so I assumed it was something else, and considered maybe you were talking about a Fuji (since base 200)
In looking at three links in the search for "100-400mm filters quality" I see nothing to suggest that it won't take a filter without fucking the image quality, except for the usual spergs who refuse to ever use any filters on any lens. CERTAINLY no sample photos to show why it would be a problem. I'm guessing you don't have any either.
>>2842843
1) You learn something new every day!
2) I don't speek greek. Is it really "eye-so" or is it "ee-so"? I feel like if I were pretending to have a greek accent, I would say "ee-so"
3) I'm still going to call it ISO because I've been saying it for 20 years, and don't want to be that pedantic asshole
4) Why the hell is it all capitalized?
>>2842842
I've never seen filters cause any issues on anything but really wide angle lenses, with vignetting. What is it about that lens that makes it special that even good filters mess with it?
>>2842835
The 5D3 doesn't have native ISO 50. It overexposes at base ISO, which is 100, and pulls it in post. You get slower shutter speeds in daylight, but you run the risk of clipping highlights and get no reduction in noise. I don't know why that guy was using ISO 50 for his dusk tripod shots, aside from an unfamiliarity with how his camera works.
ISO is a stupid nonsense name. Let's decide a better name.
>>2842310
>because they designed a sensor that gathers light more efficiently
So why is it LBCAST used a minimum of ISO 200 while the immensely better Exmor R BSI-CMOS in the A7RII still uses ISO 100?
>>2843149
subjective noise value reading
>>2843163
Different output curves in the ADC result in different ISO values per gain setting
>>2843218
So it was most likely because Nikon reused the ADC from the D1 series while the photosite area increased?
>>2843194
sensor speed
>>2843149
isi
>>2843512
I like this.
>>2843149
Electronic Gain.
>>2847261
If you don't keep the gain on 100 you're a FRUCTIDOR
