Видео

Thank you!

That's because it's not the ISO Value that makes 12800 less noisy, it's the technical/electronical things that happen once the Sony A7SIII is set to that ISO value. It's a bit tricky to explain, especially because camera manufacturers are not that open about how their Dual Gain/Dual native ISO systems work and I'm not an expert when it comes to the real technical stuff. But since the analog circuit changes at that ISO, there is a lot less noise before the ISO comes into the mix. Imagine it like this: The sensor is more sensitive to light at that ISO, therefore you can get further away from the Noise floor with less light hitting the sensor. (That's not how it actually works, it's a highly technical process, but this explanation could help you wrap your head around why it still stands that ISO comes into the mix only afterwards on digital cameras).

This can definitely be a helpful thing. Though I think it makes more sense on some cameras than others to do it.

Thank you for your kind words!

Didn't even know this was a thing yet, but sounds pretty great. I've wished a couple of times for the Sony to have a good noise & DR performance somewhere around ISO 4000-5000.

Seems like you've got quite some years under your belt. Cheers to that. I'm lucky to be old enough to have at least shot a few photos on film when I was around 5-7 years old. Also my Girlfriends Grandpa was very much into photography, he's even written a book about analog cameras for the masses. He hands us some old pinhole cameras from his collection from time to time. What you explain in your comment does make a lot of sense. However may I ask what cameras you were using this technique on? As I think this would be a workflow better suited for either film or older digital cameras, since the dynamic range and noise performance of modern RAW photography allows you to capture so much detail even in the brighter highlights and darkest shadows that you would barely need a workflow like that. (unless you're going for absolute pixel-peeping best quality).

That's indeed an interesting question and I've not tested it. What would I explained you should not be able to. However, I'm quite sure you are able to avoid clipping the blacks that why and here's my explanation for that (don't quote me on that, it's probably very wrong on the technical part and might be BS altogether 😂) The sensor captures an analog signal. This signal will always have some amount of detail, even in the darkest unless you really have 0 light. Because if the sensor would not capture the detail (or the light) initially, the ISO would not be able to make it appear brighter. The important thing to note lies in the fact the blacks or not "clipping on the analog signal", they are clipping on the digital output, which is much more limited than the analog signal. So while the analog signal gets converted to an electronic signal and then to the digital image that we see, the detail in the darkest darks might get "cut" at a low ISO, because the digital signal can not retain them. However if you shoot at a high ISO, the dark areas get brighter during the process of converting the analog signal to the digital image and therefore they can be retained.

Thank you for your comment! I really appreciate getting a bunch of constructive comments on this video that actually spark some proper conversation and not just bashing each other on the internet. I believe there's a couple ways how knowledge like this might change someone's approach. 1. (like you mentioned) while shooting video and not RAW photography, it's much more beneficial to get your settings right as your output format is not as flexible. So know things like this can make a big difference in image quality. 2. From my experience a lot of new photographers couldn't really wrap their head around ISO yet and are a bit scared of it as they think it makes everything super noisy. Some of them have not yet understood that basically more light = better image quality. Or at least they have not yet understood why that is and how you can use it to your advantages by simply lowering the shutter speed a bit instead of raising ISO to get a proper exposure for example. 3. Once you acquire enough knowledge about photography/videography you will at some point understand that a lot of the things that people keep rambling about on the internet are simply not that important (basically would you said and what I also believe). However, to get to that point you first have to understand how it all works to then understand that you can simply ignore a bunch of it (or simplify it a lot).

Muchas Gracias.

Appreciate your comment! I do believe the radio analogy really makes this a lot easier to understand for new photographers/videographers.

Do you mean base ISO? The base ISO on the T6s is 100. I don't think it has dual gain ISO. It does have an extended ISO that will let you go even higher but there's so much noise that the photos may not be usable. I have taken photos above 8000 ISO on my T6s. They were body but Lightroom was able to clean up most of it and that was several years ago. Lightroom has the new denoise feature that does an even better job cleaning noise up.

@ 100 iso is crazy

@ wow you had a t6s before what do you have now

@LouisLuzuka the T6s was my first DSLR. I got it right after it came out in 2017. I switched to Nikon and bought a d750. It got damaged at the beach and I replaced it with a z6 II. Last year I added a z8. I still have the T6s. I'm thinking about having it converted to either infra red or have it fully converted.

@ can you tell me about the Nikon series and your experience please I had a d700 and I loved that camera I still have my lenses but I was looking at the d850 cus nothing replaces that feeling of the shutter but you got the mirrorless Nikon what is your experience with them and what do you suggest thank you 😊

Thank you! Appreciate it.

How is it clickbait? The video literally explains that ISO does not create noise, it just amplifies noise that is already there. That is not the same thing. The amount of noise is always the same. Only the amount of visible noise changes. And that not only changes through ISO but also through increasing the brightness in post in the exact same way.

​so you could almost say... It changes the sensitivity.@@marcelohm123

Yes, it’s invariant except for the second native ISO or the high gain ISO. And yes, looking at it now (and some of the comments) I should have done a better job at at least mentioning the different sensor types. However I do believe that most modern mirrorless cameras are ISO invariant.

That's not what ISO invariance is. You could shoot at ISO 100 with an ISO invariant camera and raise the exposure 5 stops to 3200 in Lightroom and have the same amount of noise as shooting at ISO 3200 in camera. I don't know if any of the new Canon cameras are ISO invariant so they need to be shot at the correct ISO. A lot of Nikon and Sony cameras can be shot at much lower ISO to keep dynamic range and then raise the exposure in post with getting color noise and having the photo fall apart. You can't raise the exposure that much with a Canon camera without the photo looking like trash. So ISO invariance is a totally different subject. ISO invariance is more of a character of how some cameras ISO works.

@@carlmcneill1139 it's exactly what he does in the experiment. so it's exactly what i say it is lol it's not a huge feature or an amazing feature to have. some have it, some dont'

@Juventinos well like they say, gear doesn't matter until it matters.

Thanks for the elaborate feedback. Could you explain how ISO is not amplification? Is amplification just the wrong term or are you suggesting that the ISO setting does not “amplify” anything at all? Also I never said that the noise gets amplified more. I clearly state that the amplification effects the entire image and as you try to get to a proper brightness level by raising the ISO (or doing it in post) you also boost the noise the same amount. Never said/implied, that the signal to noise ratio changes. Regarding the shot noise: When it comes to the different “types” of noise I definitely reach the limits of my technical understanding. But from my understanding shot noise is the same for every pixel in a given exposure, but it’s more noticable in darker areas, because the SNR of the individual pixels/dark areas is worse. Also in most photos with good light you don’t need to worry about any kind of noise anyway. The explanation (especially the dual gain part) also is more geared towards Video, which in most cases is not RAW and more compressed. While shooting RAW photography I believe that most of the technicalities can be ignored due to a massive dynamic range and amount of image data captured. Just keeping in mind that getting more light into the sensor will result in a cleaner image is enough. Not taking your comment as an offense. Genuinely trying to understand if I explained something in a wrong way.

@@marcelohm123 Thanks for a very constructive response to what was intended as a constructive comment. If you're happy for a long response, I'll try to explain it all. (it might come in parts, because the RUclips comment editor is very poor). Anyway, part I as follows: First, there is a narrative common over the whole web these days which is accepted as 'the truth' by most, and survives most internet research, but which is based on some basically incorrect ideas. Generally one has to dismantle this before explaining the correct ideas. I'm not saying that you got all or any of this wrong, sometimes people will get some bits right and then try to bend other bits to fit their own narrative. This isn't at all dishonest, it's just what people do when trying to build a conceptual framework for some phenomenon - work from known facts and try to fit them together. But if the known facts are wrong, then the framework is bound to be twisted. So let's start with some definitions, and use them to work out what 'ISO' actually is, and why there is no need for 'amplification'. (End of part I)

@@marcelohm123 *Part II - definitions*. For these definitions I'll be working from the relevant ISO standards, since these definitions are in some sense unarguable. Starting with the key one - *exposure*. _time integral of illuminance on the film_ (from ISO2240, the standard defining ISO speeds for colour reversal film - 12232, the one that defines digital ISO assume that exposure is already defined) unpacking this for the less technically minded, 'time integral' means 'added together over time', so this is the illuminance (light shining onto the sensor) times the time. This is the definition of exposure is the one that has been used in photography since it was formalised in the 19th century. This is given by the formula _H=Et_, where _H_ is exposure, _E_ is the illuminance (controlled by f-number and scene luminance) and _t_ is the 'exposure time', how long the shutter stays open. *ISO* - in photography, ISO refers originally to film speed indexes, defined in ISO 2240 as: *speed * - _quantitative measure of the response of the photographic material to radiant energy for the specified conditions of exposure, processing, density measurement and analysis._ This doesn't quite fit digital usage, so in ISO12232 it was redefined: *ISO speed* - _numerical value calculated from the exposure provided at the focal plane of a DSC to produce specified DSC image signal characteristics . Note 1 to entry: The ISO speed is usually the highest exposure index value that still provides peak image quality for normal scenes. However, a DSC does not necessarily use the ISO speed value as the exposure index value when capturing images._ The lesson here is that the 'ISO' on the ISO dial of your camera is not an ISO speed, and not the same thing as 'ISO speed' was in film - important to know. Rather, it is an 'exposure index', also defined in 12232: *exposure index * _numerical value that is inversely proportional to the exposure provided to an image sensor to obtain an image Note 1 to entry: Images obtained from a DSC using a range of exposure index values will normally provide a range of image quality levels._ A quick summary before part III. 1. 'ISO speeds' define physically testable relationships between exposure and response for both film and digital. 2. The 'ISO' settings on your cameras are not speeds but exposure indexes, so there is no need for any physical mechanism to provide them, and indeed the ISO standard does not mandate any such mechanism' End of part II

*Part III - ISO in digital.* ISO in digital refers to one of two exposure indexes that the ISO standard 12232 defines. They are: *standard output sensitivity (SOS )* _specific exposure index value for a DSC that provides a still image with a specified DSC image signal value under specified test conditions._ The 'specified 'image signal value' here is: _the exposure required to produce the specified standard level DSC image signal equal to 461/1000 x Omax where Omax is the maximum output value of the digital system. For 8-bit systems, the reference level shall be 118. NOTE The code value of 118 in an 8-bit system corresponds to 18 % of the maximum final output for sRGB._ What this is saying is that the ISO setting will be based on the exposure that produces a 18% grey in the _output file_ for an18% grey object in the scene. Note that this is the processed output file, not the raw file or the 'signal' at any intermediate stage in the process. Clearly the EI ISOs don't apply to raw image, and ISO makes explicit that neither does ISO speed: _ISO speeds ... shall not be reported for raw images, however, because with raw images processing that affects the values has not been performed_ 'Value' here has a specific meaning, how dark or light the final image appears (as in the HSV colour space) - of which more later (End of Part III)

@@marcelohm123 *Part IV - words not to be confused.* Much of the confusion in this discussion arises from equivocations - using the same word to denote two different meanings, then using the two meanings interchangeably in the narrative. In this part, I'll discuss some of those equivocations. *Exposure* As in the definitions above, the correct meaning of exposure is the amount of light energy at the sensor, but often people use it to mean how light or dark the final image looks (this mistake is bolstered by mislabelled controls in tools such as LR). This error leads them to think that 'exposure' must be being increased at some stage in the process, and the usual answer to how it can be increased is 'gain' or 'amplification'. *Luminance* Even if people don't make the 'exposure' mistake, they will often use the term 'luminance', which correctly defines the brightness of light shining on something to also mean how light or dark an image looks (yet another mislabelled control in LR and others). This leads to a similar identification of a non-necessary requirement, for more 'luminance' to be provided. Again the usual answer is 'gain' or 'amplification'. *Brighter* Unlike exposure and luminance, 'brighter' is not a formally defined word. But it tends to apply to light sources, and be conceptualised that way. If you think that making the image lighter is making it 'brighter' and therefore more light needs to be created somehow. Again the easy answer is 'gain'. Note that all these three get confused with how light or dark the output image looks. To avoid getting confused we need a separate word for this. As already discussed, ISO uses the word 'value', but this isn't very explanatory. A better word is 'lightness' which is a properly defined term (en.wikipedia.org/wiki/Lightness) as -a visual perception of the luminance of an object-. That Wikipedia article is pretty good and explains how and why lightness is different to luminance. Note particularly that lightness is a _perceptual_ concept, not a physical one. No change in physical quantities is needed to provide a change in perceptual values. Incidentally, if trying not to get conceptually confused, it's best to use 'bright' only to refer real scenes and light sources, and 'light' to refer to output images. A painter adds white to paint to make it lighter, it doesn't make it brighter. (End of Part IV)

Care to elaborate?

I can't say this with 100% certainty but I think the problem lies on the manufacturers side. Color science is a highly guarded secret for every single camera-manufacturer. It what sets them apart in a world full of cameras that basically all do the same. At least slightly. And I assume without knowing how exactly a certain brands color science works it's impossible to match it 100%. And some developers might be doing a better job at it than others.

That's too technical for me. I merely understand the absolute basics of how all the technical/electronical parts play together do get from the analog input to the digital output. And I'm not even a 100% sure I understood what I understood correctly. I assume that it has something to do with how sensitive the manufacturer decides to build the sensor (so what native ISO it has) and with how the analog signal gets converted. I also know there has been a big leap when most manufacturers switched from DSLRs to Mirrorless. Though the difference is most likely due to the newer sensors that got used then. But like I said, I don't have the knowledge to answer that.

Gerne!

Yes, that's right. However going below base ISO also lowers the dynamic range. Mainly in the highlights though. So if you're looking to get the cleanest image possible and you're not in need of max dynamic range, then exposing at ISO 100 would get you the best result in theory. But from my experience there's not a lot of scenarios where this is usable advice because in outdoor scenarios where you would have enough light to expose lower then say iso 640 if that is your base, then you definitely don't wanna miss out on dynamic range in the highlights most of the time. And if you're in an actual low light scenario then you will definitely be clipping your blacks at ISO 100 or 400 for example. The only scenario I see this actually working is a semi-low light situation, where you add artificial light to give your subject a proper exposure at say iso 200 and the rest of the image which is going to be more towards the blacks is still going to be quite clean.

Hey man, I appreciate your comment! Trying to stay more consistent with it now.

@@marcelohm123Thanks for the reply. I have a request: could you please make videos more focused on video/cinematography, especially about shot continuity and framing? It would be super helpful.

I believe I spider. Glaube den kriegt man auch nur reduziert, wenn man entweder Jahrelang englische Videos macht oder ne Zeit im englischsprachigen Ausland lebt. Aber ich hab irgendwann gemerkt, dass der deutsche Akzent nur andere Deutsche "stört". Alle anderen juckt der nicht.

Well yes, if you're shooting largely compressed image-formats then you should try to get as close to your final look as possible in camera. But I'm assuming the most people don't shoot those. I might be wrong here as I'm doing a lot more video than photo these days, but I believe for RAW-Formats (stills & video) this is even more true. Videocameras that shoot RAW have completely ISO invariant cameras that have the same dynamic Range regardless of ISO settings. You just pick the ISO for better monitoring purposes but when during the editing process you get to pick your ISO in the Raw-Conversion tool and it's the same result no matter what ISO you used in camera. From what I've read during the research this is actually also true for photography. I believe the only time there's an actual difference is when your exposure was so dark that you actually did clipped the blacks. I guess that's because while the analog signal did have some detail left in the very dark areas, this detail gets lost during the digital conversion. However if you shoot at a higher iso these details get preserved during the digital conversion because the are spread more evenly across the dynamic range. (Could be talking total BS here. I'm not an expert when it comes to the extreme situations. But that's how I've gotten to understand it during my research.)

Thank you, I appreciate it! Working on the subscriber-thing 😁 This video ain’t performing too bad.

I see where you're coming from but I would disagree. From my understanding the definition of exposure in photography is actually How much light hits the sensor/film. So by definition we actually do care about the part that stops at the sensor/film. Yes, with digital imaging this definition becomes a little more loose as one could raise the question "When is an image really finished? When it comes out of the camera, or when I finished editing?" Now with that being the case you could definitely say that ISO affects the exposure, as the Base ISO setting of the sensor of course plays a role here. The reason I said that ISO does not affect the exposure is because changing the ISO does not change how much light hits the sensor. By your logic raising the brightness in Lightroom would also change the exposure of the image. And if I was talking about editing I would definitely say "Now I'm going to raise the exposure a bit", but I guess we can agree that the editing process happens AFTER the exposure has been made. So it does not affect the actual exposure itself. It changes the resulting image. Also in the big picture of the entire Video, isn't this wording just a technicality? It does not change any of the other things I said in the video and it's also not the main focus of the video. Yes, it would have been more fool proof if I kept saying "Iso does not change how much light hits your sensor." But for me that was the same thing as not changing the exposure.

That's an interesting approach! I know that for Lady Bird, they filmed a blank background at higher ISO in order to to have an asset of their specific camera's noise. Then they could overlay it on other footage for consistency sake. Perhaps this is a common practice, but I hadn't heard of it before

You think so? I've personally never tried it in a proper daylight situation. Every time I get noise issues in lower light situations I find the noise pattern rather unpleasant. But maybe it feels different in brighter situations.

A Helios 44-2 58mm, a MIR-1b 37mm and a 17mm Tokina wide angle. However lately I've really enjoyed shooting on the Tamron 35-150. Sony's Slog3 in combination with current lenses and good grading also creates a very nice look. It's a little less cinematic than the vintage lenses, but it does look very polished which I also kinda like for commercial jobs.

From my understanding there a 3 "types of noise" that make up the noise in any given image. 1. Thermal noise which is caused by the heat of the electronic parts of the camera. 2. Read noise which comes from reading/converting the electrical charge into digital data. (This is the most "defining" one for our standard exposure and the noise floor) 3. The shot noise you referred to, which is due to the randomness of light hitting the sensor. This gets much more visible during extremely low exposures (so at very high ISOs) and happens because there are a lot less photons actually hitting the sensor which means the randomness gets more visible. The term noise floor in general is used to describe the sum of all the unwanted signals within a measuring system. So the noise is basically everything else than what we're actually trying to measure. And the baseline noise is the noise floor. But I'm not a technical export. So there definitely might be some wrong wordings or stuff like that. I'm just trying to explain the things in a way that is understandable for everyone.

@@marcelohm123 Thanks a lot for taking the time to clarify, and I are really doing an excellent job explaining the things in a way that is understandable for everyone. Your summary of the three types of noise is brilliant. My comment was just to address a slight misconception that is mostly based on outdated technology. To put things in perspective, with modern sensors, read noise is so low that even astro-photographers can usually ignore it. Similarly, even though astro-photographers can get noticeably less thermal noise with cooled cameras, that's only because exposures are typically more than 100 seconds (thermal and dark current noises are proportional to exposure time). Basically, for most situations, the only noise worth considering is shot noise. The reason why I am pointing it out is because manufacturers claims on this topic can be very misleading and push us into wrong purchasing decisions.

Thank you.

Thank you!

That's true of course. Getting into the different kinds of "ISOs", different native ISOs, older Sensors and stuff would have probably made the video 20 minutes long, so I opted to leave it out and said "most cameras these days" and mentioned that the ISO Settings I'm talking about were specific to my Sony.

@ not critiquing what you said. I love the way you explained it. At one point the way I heard it could be confusing to some so just wanted to clarify.

@@cap2redstudio Yeah, not taking any offense. You're right to point it out in the comments. Might have been smart to at least at one sentence explaining that the native ISO differs between Cameras/Sensors.

@@marcelohm123 A sensor has no ISO. 'Native' ISO doesn't mean very much. There are plenty of examples of cameras using the same sensor with different 'native ISOs'.

@@BobN54 I guess it's not the sensor but the sensor + all the other parts involved in capturing the light and getting it from analog to digital that dictate the native ISO?

Thank you. You’re welcome :)

Thank you. Appreciate your Feedback.

Thank you. Yes, if your subject is a person this might be something you wanna do.