Do full frame cameras indeed have lower noise?

[@simon_dentremont]

What’s your experience with noise compared to sensor size?

[@FlatWaterFilms]

I'm thinking many people feel smaller sensor cameras are just as good as FF. Maybe a justification for them due to the affordability? The OM-1 looks pretty nice for wildlife photography though. Personally, I'm willing to lug around the R5 with battery pack while out in the forests. Glass is important too, with FF normally getting the highest quality lens. Crop sensors have lot's of reach, but going wide is a problem. In the old day's 800 ISO was pushing it. Modern cameras, no problem. Heard with the R5, if you can't shoot at ISO 100, go directly to ISO 400 for the best DR. Simon actually provides excellent information, while so many others just push gear.

[@cityproofdad]

its more about focus sharpness

[@Eikenhorst]

I think noise becomes less and less of an issue with AI tools being able to remove noise now without removing the details in your photo, unlike previously where software smudges out the noise, but also the detail. There are still advantages to going for a FF setup in many situations, but noise becomes less of a factor in that decision.

[@FlatWaterFilms]

@@Eikenhorst Not a good idea to reduce noise in post production in my opinion.

[@RG-rm9jt]

Having shot both Fuji and Sony, the Fuji did cause more noise but that's because of their lens selection. At the time they had a 400mm f5.6, so equivalent to a ~600mm f8, which of course will cause more noise due to the small aperture. To be frank though, I never worry about noise, that's what Topaz is for these days.

[@phrozenoddity995]

I am a brand new photographer. The photos i have taken so far are well above average because of everything I've learned from you in these videos. Your content is master class quality. Thank you so much for being a teacher, role model and inspiration to so many up and coming photographers!

[@oli8200]

I feel like squabbles over the relationship between sensor size and noise are almost redundant for wildlife photographers because we're just so used to what others might think of as "high" ISOs. If I can get down to 1600 ISO I'm ecstatic, but my landscape and portrait photographer buddies are horrified by anything over 400!

[@simon_dentremont]

Agree.

[@robertleeimages]

I do nightscape, landscape, day & night street, some sport and now and then like last Friday i do a bit of bird photography but only ever in our front garden(Live on a farm) But because of the nightscape etc i am use to noise so except for the nightscapes and landscape where I'm in full manual, everything else my camera(200d) is set to auto iso max 1600 and shutter priority

[@RG-rm9jt]

Agreed man. ISO 12800 is not too uncommon for me, and I find that smoothing the noise isn't too bad with modern software and sensors.

[@umfilmmaker8253]

Very good point 👍

[@SekiLapse]

Hahahaha, that's so true!

[@NotAnotherChannel_Channel]

Ah, the argument that will never die. I’ve been told time and time again that I can’t shoot the astrophotography or night photography that I’ve been taking with my APS-C camera. And those wedding and concert photographers that I follow who also shoot APS-C that they’re not supposed to be able to shoot but still seem to nail it. Modern sensors, lenses, and tech help ameliorate noise in almost any image. Not perfect, but like you said, understanding your equipment’s limitations and taking steps to reduce noise is how we get it done. Thanks Simon!

[@robertleeimages]

Hahaha I get questioned on social media(not so much now)about how could I get shots like my profile with a little 200d, and that they're fake and photoshop etc etc. I don't even use or own a tracker, photoshop or lightroom and still edit everything in Canon DPP4 before stacking sky images in sequator, then it gets combined with any light painted foregrounds using layers in Gimp. Canon 200d with Tokina 14-20mm f2 lens, that's all my nightscape kit is

[@NotAnotherChannel_Channel]

@@robertleeimages Weird how people were doing photography before 2020. And much respect to you!

[@Warrior_Resisting_Colonialism]

It's not that you CAN'T, is that's you can do it FAR better using a full frame camera. I got into photography for shooting the Milky Way myself. I started with a d3400 crop sensor, then upgraded to a used full frame, D610 and the image quality FAR surpassed anything I could take with the d3400 by a LARGE margin. So far superior, it wasn't even like comparing apples to apples. The VERY first thing I noticed, was once in lightroom, I could use a MASSIVE amount more of all the sliders!!! Meaning, the D610 captured FAR more in the raw files. Where if I slide any slider so ever so slightly on the d3400's raw files, the imagine went to crap real quick. Someone explained to me, it's because the larger sensor captured more "dynamic range" than the crop. I wish I had known there was no comparison, I would have just skipped the crop sensor camera. It was a lesson that cost me a LOT of time and a LOT more effort to get decent night shots. With the D610, it's feels like I entered a god mode cheat code in a video game or something. There's a right tool for the job, and then there's making due.

[@alansach8437]

Nothing motivates better than someone telling you that you can't do something!

[@Warrior_Resisting_Colonialism]

@@alansach8437 ALSO, nothing educates better than ensuring when you do use a crop sensor camera for night photography, that although it will work, it wont work anywhere near as good as full frame camera. This way people will know that they CAN do something and they can do it even BETTER. This way the choice is their own. No ones wrong either way, but we can always ensure to be informative and not be misleading by leaving out vital info. I question that anyone actually told the first poster that he "couldn't". I've taken plenty of decent night photos with a crop sensor and it's very will known you "can". Only very uneducated would have told he "couldn't". It's important we all know a full frame can do this task far better, that's all. Not that you "cant" do it. This video explains WHY but it's tip toeing around that full frames are FAR superior in low light over smaller sensors. The difference is very, very drastic.

[@coolcat23]

Few people do this subject justice, but you did. Well done. A crucial point to understand, that many don't get, is that "the same exposure" only refers to the same intensity, i. e., number of photons per unit square, and that overall IQ depends on the overall number of photons captured (as opposed to local intensity).

[@gregsullivan7408]

Agreed, and this is a point which Tony Northrup explains very well.

[@VickiTraud]

I love the way you explain things in simple, user-friendly terms! Thank you for sharing your advice and knowledge with us Simon! So much appreciated!

[@simon_dentremont]

Glad you like them!

[@kenmaier6870]

This is by far the best / most thorough explanation. Your analogies are great !!

[@simon_dentremont]

Glad you think so!

[@12345678927164]

This guy always states the facts and makes me more confident

[@MurrayVader-xp8iv]

I appreciate your last tip that helps people to get the best out of what they have. I'm not in a position to upgrade.

[@joeoneill9098]

In think a better way to tell how good a sensor is at gathering light is seeing how much ISO the camera wants to properly expose the image. Restrict the shutter speed and aperture and see what the camera wants to do with ISO. If one camera can have a lower ISO to achieve exposure, and the other needs a higher ISO, that would tell you that the lower ISO camera is doing a more effective job of gathering light. I don't think whacking ISO up on normal pictures is a true test of low light performance for cameras. All it does it tell you how much grain will be introduced

[@simon_dentremont]

The « effectiveness » measure you raise is called quantum efficiency, the percentage of photons that are actually captured. Camera manufacturers sometimes publish this. but your experiment wouldn’t work, as the camera’s iso is normalized to a standard brightness, so the iso would be the same in both scenarios (but the better quantum efficiency sensor would have a cleaner image, everything else being equal).

[@G95G95]

A sensor is like a solar panel, bigger creates more output, not complicated.

[@Warrior_Resisting_Colonialism]

Close enough. That's really all we need to know as photographers. The important thing is that we all know it to be true and not get so lost as to how, or else it confuses people into thinking their crop sensor is just as capable as a full frame and it's not.

[@mightygame9944]

This was the perfect explanation that i was looking for, compare to many youtubers that couldn't explain the proper reasons on low light between full frame and crop sensors.

[@falxonPSN]

The newest technology of sensors, especially when paired with a low megapixel count for a given size, can produce some amazing results. I've been testing a technique on my R3 when taking pictures in near darkness that consists of using video shot at ISO levels of up to 51, 200 and then stacking frames to reduce noise. It works amazingly well and it can basically see in total darkness. People love to say that the R3 stinks because of it's low megapixel count compared to the R5, but when it comes to low light there's no comparison.

[@Aranimda]

I upgraded from a 1/2.3" sensor bridge camera to an 1" sensor bridge camera. While this is still a very small sensor in the eye of professional photographers, it made a big difference to me. With the versatility of giving a lot of sharpness and versatility when zooming.

[@kedrednael]

Let me guess. From Sony HX400v to RX10 IV? I am going to try to do that transition next week.

[@Aranimda]

@@kedrednael From a Canon SX20 IS to a Sony RX10 III.

[@davidcrossley7145]

What a very informative photographer to learn from. This man has a great way of explaining difficult and awkward problems I face regularly in my photography.So thanks so much for your videos 👍👍

[@simon_dentremont]

Glad it was helpful!

[@drtod]

I simply use DxO to clean up the noise, its a great equalizer against sensor size.

[@JohnDoe-xm1ir]

I haven't tried DxO, but Lightroom's AI denoise has been absolutely fantastic for me personally. It's slow, but it can really revive photos that just had a bit too much noise.

[@dogpadogpa]

Great video! I've debated in your vids when people say FF is always better than crop. However it's very difficult to compare sensor size on varying technology. What I do to explain easier is to compare quarter size to full size on film. Comparing the same film like Kodak Ektar100. Full frame can print a 4 times larger page than the quarter frame. Both grains look the same when looking at the same distance. Full frame looks less noisy when printed at the same size prints. Full frame cropped to a quarter is exactly the same on all levels to the quarter frame camera (IF your lens can resolve that detail). Also why panoramic stitching and high res stitching helps in smaller sensors too. AKA full frame is good for larger prints. If you don't need larger prints (because viewing distance matters like you said in your previous video!) then crop sensors are fine for MOST people.

[@H0mework]

Noise is now an artistic choice as well. There's been a trend of "filmic" digital with CCD sensors recently. Love your videos and productive.

[@blindspotter5859]

Your explanation is simply amazing...wow.

[@rocheuro]

pixel size is crucial, not necessarily sensor size. but often one determines the other.

[@IllusionInfusion]

My philosophy has always been that a picture with noise isn't great, but a blurry picture is useless. Obviously lower ISO is ideal, but too long of an exposure can kill a shot more than noise can.

[@simon_dentremont]

Agree!

[@JezdziecBezNicka]

Bigger sensor is better, but there's an individual cut-off point called "good enough". Once you reach that, further investment gives massively diminishing returns. For example, if the only place you display your photos is on social media, getting a medium format camera makes absolutely no sense.

[@simon_dentremont]

Agree!

[@josgeusens4637]

Digital sensors - by the way - do not stop at medium format. Large sensor cameras use 4 by 5 inch (10 x 12.7 cm) huge ones. Just fyi. On topic, I might add that some people don't use the same lenses to compare full frame and micro-four thirds because f/2.8 is not the same as a 2.8 T-stop. While this is marginally important, it's also true that comparing images on pixel-level is marginally important. Very often it's also ignored that ISO 200 is not the same for all sensors, not even for the same sensor size and brand (1). This especially shows up when using a reliable external light meter (Sekonic L-389) and setting the camera to that readout. (I never do that, btw, because the most reliable meter is built in the digital camera itself for it measures the light the sensor receives and no deviations are included). To reduce the noise, we can also use the very high frame rate of the Olympus cameras and stack 12800 ISO images. While at 120 fps there is very little differentiation between the images, this can be a helpful feature. (1) Sensors are not always manufactured by the camera manufacturer. Mostly not, except for Canon and Sony.

[@mitchellan-ebbott7408]

Great video. One advantage to smaller sensors that you didn’t mention is stabilization. Smaller sensors with less mass are easier to stabilize, which is why the best IBIS on the market is on micro four thirds bodies. For stationary subjects, this can mitigate the poor noise performance because it lets you use a slower shutter speed.

[@proksalevente]

The Sony a7rV has 8 stops of IBIS. Same as m43 flagships. Canon also has 7-8 stops already iirc.

[@r4yker442]

Isn't it true that with FF gear, which in most cases is heavier, introduces more camera shake that IBIS has to battle? MFT not only is rather compact compared to FF, but in most cases lighter as well all while having excellent IBIS

[@PhreddCrintt]

Spot on Simon. Again. I have been saying this for decades. (I shoot Olympus E-M1 and Canon 5D Mk3 - both with OEM glass)

[@Lesterandsons]

Excellent. FF requires longer focals, so larger apertures at the same f/ stop to cover a wider area Same exposure but more light. To take advantage of FF the size of the subject has to be larger on the sensor. If the size is the same it’s a waste. Apsc are today very good and software makes wonders.

[@davidgommeren7283]

Thanks so much, the comparison with rainfall makes it very easy to understand what is going on.

[@johngregg5735]

I took your advice before I even saw this video. My DSLR is one of the best C2K cameras around. C2K as in Costco Camera Kit. A Nikon D3200 w/ a 18-55 ( f/3.5- d/4.6 4) and a 55-200 (f/4-f/5.6). And it came with a real nice bag. I've been trying to take photographs in low light, without a flash. Results were not so good. Yesterday, my (relatively) new 50mm f/1.8 was delivered. Just messing around, I was able to take photos in very dim light and actually get good images. Of course, they put the 'G' in grainy. A game changer.

[@timgurr1876]

Thanks Simon. Great explanation. I keep using APSC because of cost. In low light I use wider apertures to reduce noise (aperture mode on my Sony a6500). Also, I try to keep ISO lower than 800 (the lowest possible to keep a reasonably fast shutter speed (1/60 or faster for hand held).

[@LeoS-58]

That was a great explanation of the noise issue! Thank you Simon.

[@MasticinaAkicta]

Having a full frame saved my bacon. Though, its size means it doesn't goes everywhere with me. Envision, a small event and you are asked by family to help out with some pictures. For the local news paper, the website and organization. Since I didn't want to travel super heavy I went, one zoom lens, full frame camera model, and just hoped it was enough. Light conditions? BAD! NO FLASH! Camera doing hard work, fluctuating between ISO 2000 and ISO 6400. That is the situation. In the end though while some pictures were unusable enough were usable. That big sensor definitely saved the day there. I did have a 1" camera with me but... looking at the light situation I just knew it would be unable to do anything for me.

[@JRodPhotoArt]

Well done Simon !! Everyone should watch this and maybe people will stop complaining.

[@MeAMuse]

Very well explained! You obviously have to leave a bunch of the complexity out. The way I explain it is that larger sensors, and better lenses make it easier to get a photo (they don't necessarily make the photos better). I mainly shoot full frame, but even when I shoot 1 inch.... I can get great results by good technique and working the light.

[@ion_X]

Great in depth comparison, I do love these technical analysis

[@RG-rm9jt]

As someone who shoots wildlife with a 600mm f6.3, noise does not bother me in the slightest. I use Topaz for that, and will happily push my iso up to 12,800 if needed to get the shot sharp.

[@dankedozo]

This is such a great explanation! I really appreciate how you clearly define terms in your videos!

[@simon_dentremont]

Glad you think so!

[@chrisburnard5157]

Thank you for the hard work you must have put in to make this content.

[@simon_dentremont]

You bet

[@mrv1264]

At 4:20, the larger sensor does capture more light (more absolute amount of light). At 4:30, the smaller sensor __might__ capture the same amount of light per area (not the same absolute amount of light). Whether it does indeed capture the same amount of light per area (density) is a function of the shot noise and read noise performance of the camera that you mentioned at the beginning of the video. At 4:50, you should say the smaller sensor captured the same density of light (not intensity), density being defined as photons per unit area.

[@rudigerwolf9626]

Simon, thank-you for the time you take to address multiple issue and bring some reason to the many controversies in photography. Your approach is the most reasoned I have seen on "KZbin"! Thank-you! I just watched the recent video on Larger Sensor, Lower Noise. I truly appreciate the explanations. I would suggest looking at sensor size without real world camera/lens capabilities is an incomplete comparison. I happen to have rented both a Nikon D9/800mm f/6.3 and OM1/150-400mm f/4.5 for testing. At an effective 800mm they have the same field of view. The cameras have way different resolutions. Lenses have different minimum apertures. Testing both lens/camera combinations on a tripod, the OM1/150-400 consistently delivers clearer, sharper images when viewed at 200%. At normal fit, the two are very similar - except I can typically shoot the OM1 slightly faster or at lower ISO due to the faster lens. Shooting at lower ISO typically results in lower noise. My point is, we cannot compare sensor size noise independent of real world lens and camera capabilities. I am not saying one is better than the other. What I would personally love to see is real world comparisons where a number of parameters are assessed for different shooting genres. Parameters would include weight, size, cost, sensor resolution, stabilization, lens focal length and aperture, noise levels, simplicity of use, ergonomics, lens portfolio, tracking, frames per second, minimum focus distance etc. Genres might be landscape, sports, wildlife, portrait, micro, astro, street, hiking landscape, etc. Finding the right tools for the job (systems, camera/lens combinations for a high quality 11x14 print, or instagram post, or online display) should be the point of these types of videos - at least for me. I have yet to find someone who approaches the reviews in that context. Perhaps it is just too time consuming? Finally, my thanks again for your videos. I truly enjoy your reasoned approach.

[@dogpadogpa]

I followed that path of comparing cameras (by looking at all those parameters including weather sealing). I ended up with m43 as my go to (still playing with medium format/135 film SLRs). The unfortunate thing is bigger sensors and high megapixels is what people get sold by. Easy for marketing and profits for the company. Each system has their own strengths and weaknesses (some full frame fans don't understand that even full frame has weaknesses). Having higher megapixels won't help you if your lens can't resolve that detail. If your camera system makes you enjoy photography that's a good system for you!

[@rudigerwolf9626]

@@dogpadogpa Completely agree. Most of us likely won't be selling any large volume of prints. So the joy of traveling, camaraderie, and the pleasure of taking the shot is where the joy is. So the equipment that makes the experience more enjoyable is a key aspect of equipment selection. Enjoying the finished image is another aspect. Personally, I am right there with you. Medium format for landscape, nature, scenes, portraits. M4/3 for action, wildlife, macro and light carry. Still have a full frame and APSC, but not really using them very much.

[@dogpadogpa]

@@rudigerwolf9626 the rate of technology has also helped cropped sensors. Back in the day I had to rent full frames to do low-light action shots but now m43 is fine. Printing is fun but it's funny how dynamic range has a big concern when paper's dynamic range is the smallest range ever! Well, even general phone and computer screens can't show off dynamic range (it's mainly for having lost detail to highlights and shadows). I still prefer full frame for bokeh and portraits, tilt-shift lens for architecture but current m43 is fine for everything else.

[@rudigerwolf9626]

@@dogpadogpaEspecially travel. OM1 and 12-100 is an awesome travel combination.

[@dogpadogpa]

@@rudigerwolf9626 Agreed. Or the 12-40 or Panasonic's 12-60. There's many choices.

[@adunthecitadel9122]

What about thermal noise? Lights are photons received by receivers through antennas. Thermal noise in receivers make noise goo astrophotographic cameras lower that heat reduce your noise. Does smaller sensors generate more internal heat? Is there interferences between pixels? In AstroPhoto, Fullframe requires costier optics while the small sensor fit with much more telescopes.

[@simon_dentremont]

Thermal noise from hot electronics is only an issue for long or repeated exposures (ie over 30 seconds or timelapses). I own cooled astrophotography cameras for this purpose. For daytime photography, just taking a photo here and there, it’s not an issue.

[@deejayiwan7]

Soooooo you are right... Thats why everyone needs a Flash... Or two... Or three....

[@thatcherfreeman]

I love the continuation of the catching rain in cups analogy. Super intuitive explanation!

[@MatthewUseda]

Finally you mention medium format. I'm happy now! Love your videos

[@simon_dentremont]

Assuming you have one!

[@MatthewUseda]

@simon_dentremont yes! Can't wait for you to test out the GFX100 II with the new phase autofocus and more FPS. Maybe! Have fujifilm have you do a video on it. They also will have a 500mm lens coming out to go with it

[@samue1991]

Your experienced and well researched perspective is a blessing in the photography community, thank you for making these videos!

[@simon_dentremont]

My pleasure!

[@careylymanjones]

I usually shoot in manual mode, with auto ISO. This lets me control my depth of field with the aperture, the amount of motion blur with shutter speed, and lets the camera figure where to set the ISO,

[@simon_dentremont]

Same!

[@haydenlee8332]

Amazing video Simon! I am especially annoyed by "equivalence" arguments because I too think they are very short-sighted and simple-minded arguments. As you've said in this video, reaching the "true equivalency" is virtually impossible: 1) *Field of View.* This one is the easier one to overcome. In order to fit the same amount of scenery into a different sized sensor, the industry has been putting out equivalent field of view lenses (e.g. 24-70mm for full frame, 17-50mm for aps-c, 12-35mm for m4/3) for decades. 2) *Bokeh.* This one is the problematic factor. We can easily see opinions online saying stuff like "in order to emulate the full-frame's f/2.8 bokeh, m4/3 will need something like f/1.4." As you've said in the video, while f/2.8 28-75mm lenses do exist, f/1.4 12-40mm does not. 3) *Focal Depth.* This one seems to be very overlooked and becomes especially problematic if paired with bokeh. Let's get back to the example from #2: Let's suppose we somehow magically emulated the blurriness of full-frame f/2.8 with microfourthirds f/1.4. You probably would have noticed that the depth of the image that is in focus is shallower on the m4/3 than the full-frame (e.g. on m4/3 only the eyes of the subject may be in focus, while for full-frame, the subject's entire face is in focus). This is pretty much similar with the issue with noise, because the larger sensor size allows us to have longer depth in focus while still having similar amount of background blur

[@simon_dentremont]

Agree! And every different genre prioritizes a different meaning to equivalent.

[@rominnooo]

I love those videos, you are one of the fewest people that can explain everything so good.

[@simon_dentremont]

Glad you think so!

[@runcmd1419]

More light over more area, but assuming the same noise performance per area, the smaller sensor needs to be enlarged more for final viewing. Same as with film.

[@simon_dentremont]

Agree

[@AnandaGarden]

Thanks for understanding the needs of your viewers and helping with such clear explanations. I am grateful.

[@simon_dentremont]

You are very welcome

[@neilcousineau4956]

Thanks, I am staying with MFT - light weight, low cost lenses make it a fantastic everyday - everywhere carry camera.

[@simon_dentremont]

Good choice!

[@dhrubo9140]

Tnx ❤ we want more videos on aps c sensor camera

[@simon_dentremont]

I will try my best

[@flatheadprints]

Your knowledge is beyond phenomenal, well done.

[@FilmRepair]

Shout out to keeping MFT alive

[@enzocannizzo7411]

Thank you for your excellent presentation on sensor sizes .As a M43 photographer I understand that due to laws of physics there are limitations w/ M43 .However a camera is a creative tool .The person who employs the tool must understand what are the limits and parameters of the successful use of the tool to accomplish the task . In conclusion there is too much chasing the tail w/ gear .Thank you I look forward to your presentations as mini lectures !!!

[@simon_dentremont]

Glad it was helpful!

[@miker5502]

Simon has some of the best and informative videos on photography in all of KZbin, all presented in an interesting and understandable way. Thanks for sharing your knowledge with us. Cheers MikeR.

[@Tugela60]

The level of noise is determined by pixel size and dynamic range of the associated electronics, not sensor size.

[@mikejoyner3051]

hmmm not all of it, pixel size also determines row and column repeat pitch, and signal integrity is also affected by this, especially column-to-column.

[@colinblin1723]

This was a very informative video. I especially liked the part where you explain the arguments online and the analogy of cups collecting water to the sensor collecting light.

[@bigrobotnewstoday1436]

With DXO Deep PRIME on my Olympus OMD E-M1ii at ISO 3,200 the image is 100%. I've seen full frame cameras at ISO 25,600 with DXO Deep PRIME 100% clean.

[@rocheuro]

interesting things happens when you use a "Speedbooster" adapter/with 0.71x with FF lens on mirrorless APS-C camera. this becomes really interesting.

[@ww8wv1]

Great explanation, I have both a crop sensor and full frame mirrorless cameras. What I’ve noticed is that in good light the crop sensor can stand toe to toe with the full frame but in lower light the full frame pulls way ahead. If I’m understanding this video correctly I now have a better understanding as to why this is so.

[@simon_dentremont]

Yes, exactly

[@Warrior_Resisting_Colonialism]

Yup. I wish I had known from the start just how drastically the better the full frame was over crop, since I was interested in Milky Way, nightscapes. I had to learn the hard way haha. I feel I was wasting so much time and working too hard, trying to push the crop beyond it's capabilities. All I ever saw were these video fights as to why, but no one said how drastically better the full frame actually performs!

[@BuildingByFaith]

Great video. I'm a m43 shooter and am thankful for modern AI noise reduction. My question for crop sensor lenses is why not project a smaller image circle? Then the light would be more concentrated onto the sensor.

[@simon_dentremont]

Wider apertures and sorter focal lengths can indeed concentrate more light and a smaller image. So settings and gear choices are what controls those. But then in comparison, the larger sensor will also get a better image.

[@tysonator5433]

Here is my take on this, the ISO noise on my 7D got bad after 800, my 7Dii it was apx 3500 - 4000, my R7 I have got reasonable results from 6400 and even 10,000 ISO ! My 7D images at this level would be completely unusable, and though R7 users complain about the noise it is still very, very good image compared to 5 years ago. Plus we have a lot of DeNoise software to use. Any camera will work amazingly in good light, however we do not also have that luxury all the time, so we need to master the techniques to achieve better results, not rely on the camera to solve the issue for us !

[@paulengle5784]

Holy s**t, photons as glasses during a rainfall is a fantastic way to explain it.

[@DannyTaddei]

I’ll start by saying I love your videos. Thank you. So I’ve been loving my Canon M6 mkii with a speed booster and EF lenses. I have two of them and use them primarily for local store social media commercials and music videos. My 50mm f1.4 comes in with an equivalent f0.97. I love the results, lightness of the camera, everything. It’s seems to me to do magnificently in low light. I’ve never done the testing like you do or other do. All I do is love the result and ease of use… oh yeah, and the extra profit I put in my pocket by used a $1000 body instead of a $3500 body.

[@brentfugett2700]

Excellent dive as always Simon! I agree, the forums are ripe with people making a big deal out of inconsequencials. I think the one addage I've heard that i agree with, to the point, don't get hung up on sensor size: date the body, marry the lens. Ie invest more in glass than camera bodies which tend to come and go.

[@simon_dentremont]

Agree!

[@ion_X]

So basically light and noise ratio per surface are is very similar, if pixel size; and once again if sensor tech and type and it's photosensitivity are similar

[@danceswithferrets]

I grew up on film cameras, I find it counterintuitive to mess with the ISO. I'm learning new skills and make more use of digital camera settings than I used to. I have an Olympus m4/3 camera and a couple of very nice lenses. Although with the pancake kit lens I can take it anywhere and get the pictures I would have missed. A small and light camera suits me best. I take the best pics I can with it and occasionally surprise myself. If my camera has a noisy sensor so be it, I really like some of the pictures I take with it.

[@juliette-mansour]

Hello Simon, I'm new here and so happy to have found your channel! I shoot street photography and have been shooting for a couple of decades. Only the last two years have I become interested in nature photography. I bought a Fuji XH2S because of its versatility in both genres and improved AF. However, I'm barely hobbling on with the 70-300mm and the lens selection is killing me! My specific situation is that I cannot carry heavy zooms (I'm 5'2" with tiny hands and suffered on and off with carpal tunnel/other hand issues). I can stand to use a camera all day as long as it weighs a total of 4lbs or so and less. My preference also is to hand hold (no tripod) though I do bend sometimes. I don't really want to invest in micro four thirds for birding, but will if that's my only option. Right now, the XH2S or something from Sony (but so far even with the smaller body, I'm not sure if there's a lighter weight lens that would serve for bird photography) or the Canon R7 are the options I have looked at. I have several Fuji lenses (mostly for street) and a few Nikon F mount/DX mount lenses. I'd prefer Nikon but the Z system is insanely expensive and I'm trying to keep costs down. I know that sounds like a lot of restrictions, but would love it if you have any time at all to provide a little guidance as to go for the OM-1, keep the XH2S and keep using what I have or sell that and go for an older Nikon or Canon DSLR? I know there must be an option that makes sense. Thank you!

[@simon_dentremont]

A light long lens is the issue, rather than the body I’d say. The Canon 600mm f11 on a Canon R8 would be light and solid autofocus. As long as f11 isn’t a deal breaker.

[@juliette-mansour]

@@simon_dentremont I appreciate you taking the time to read and respond. I will look into it!

[@tomholzwurm86]

Thanks to Simon for the valuable summary and I vote for the full frame sensor ! I also agree on permanent improvements by AI tools, I've included AI denoising in my Lightroom import workflow permanently, the only downside is the time for processing the raw images. On the brightside I'm able to shoot most of my images using a Nikon Z8 in manual mode, means only ISO is automatic. Shutterspeed is set according to avoid motion blur and aperture to adjust depth of field. I was on a holiday shooting lots of night images and looking at the statistics of the images many of them were taken using ISO 3200-12800. After processing most of them are perfectly sharp without visible noise. I also like the possibility to switch from FX to DX mode especially if even the 600mm tele lens is too short. I fully understand the need for a small more portable camera systeme, but I got used to the FF camera and the size and weight of the equipment and I'm very satisified with it.

[@joylox]

I had an image at 40,000 ISO, and used Topaz Denoise combined with a bit of extra tweaking in Darktable, and the faces looked a bit weird and blurred, but overall, wasn't bad. And if you're only viewing it at 25% of the full image quality (like using a 1080p screen to show a 4K image), it's not bad. There is still the loss of detail, but some of that can be filled in again if you know what you're doing.

[@washingtonradio]

I think the more one understands sensor physics the better able one is to compensate for the sensor size by using different camera settings, different lens choices, and general technique in more challenging conditions. Also, one needs to be very familiar with one's gear to get the most out of it.

[@simon_dentremont]

Yes, exactly

[@simval84]

The night shot mode used by many cameras is basically a way to use a form of electronic stabilization in still photography. With a camera with IBIS, you can take a photo with a long exposure handheld, say 1/10 s at ISO 640, which you can't without some lens stabilization which is limited to 1/FF equivalent focal length (say 1/25 s for 24 mm). By taking say 6 photos at 1/60 s at ISO 3200 and using computing to make each photo fit and average the noise, you produce the equivalent of a single 1/10 s photo at ISO 500, more or less.

[@David_Quinn_Photography]

I use crop sensors and never heard of such a thing but I did like your analigy with rain drops on glasses.

[@christophhoppe2947]

Simon, it’s always a pleasure to watch your videos, thanks!

[@simon_dentremont]

Glad you like them!

[@wandererstraining]

Great video, your explanations were very clear and I hope that it helps people understand better how it works and various tradeoffs. Too many people don't get the difference between intensity of light vs total light captured by a sensor. It's always about finding the right combination of sensors and lenses. A 135mm f/1.8 give an APS-C sensor the equivalent of a 200mm f/2.8 on a full-frame. Most people don't have access to a 200mm f/2 lens, so one should compare the performance of both equivalent lenses. In this case, a good 135mm f/1.8 on an APS-C sensor might give a better performance than a 200mm f/2.8 on a full-frame sensor if the zoom lens isn't as sharp or has worse vignetting or coatings than the f/1.8 lens. All in all, the two would be pretty comparable given the same sensor technology. The APS-C camera might have an easier time to focus in low light with the 135mm prime. In general tho, full-frame cameras win the equivalency game. It would take an f/0.6 lens to gather the same amount of light on a m43 camera as it does on an f/1.2 full-frame camera. For APS-C, it would take an f/0.8. No f/0.6 lens was ever made, and f/0.8 lenses would be incredibly difficult to correct and their image quality would never compare favourably to the full-frame equivalent, just like current f/0.95 lenses for APS-C cannot match an f/1.4 lens' quality on full-frame. The only f/0.95 lens that would be an exception to that would be Nikon's Noct, which weights a lot, is extremely expensive and only has manual focus. Also, it's a full-frame lens anyway.

[@wandererstraining]

The sneaky trick that smartphones can do to reduce noise, you can do with a camera as well if you're willing to do some post-processing. The trick is to take a burst, put it on the computer, align all the frames, and then apply median blending. It can be done in Photoshop, or with open source tools like the "align_image_stack" and "convert -evaluate-sequence median" commands in Linux. A cool thing about this is that you can get better dynamic range out of your image files. Process the raw files almost as heavily as you want, the the noise that shows in the shadows will be attenuated once you blend the files, possibly leaving only sensor read noise if your sensor is dated. The amount of noise reduction works the same way as f-stops. 2 images is one stop, 4 images 2 stops, 8 images 3 stops, 16 images 4 stops, etc.

[@neurologistFACP]

Superb, as usual. The content is not new to me, but the way of delivering the information is quite simple and viewer-friendly. Keep up the good work!

[@simon_dentremont]

Great to hear!

[@sdm2]

Recently discovered your channel. I use both MFT and FF and have researched many, many videos on this topic. You explained this perfectly!

[@simon_dentremont]

Welcome aboard!

[@malikknows3510]

Very well done, thank you. I’m a M43 shooter because size, weight and cost was important. I also have Topaz AI to reduce noise and even the playing field a bit with larger formats. At about $200, it is a cheap upgrade to a smaller format system that allows me to get results closer to larger, more expensive systems. All the best.

[@enzocannizzo7411]

What is your take on. TOPAZ AI in its application w/ M43 gear .For the record I just purchased Topaz AI need to start working with it ..Thanks

[@malikknows3510]

@@enzocannizzo7411 I think it is a gamer changer. It's still early days and it will only get better. My guess is there will always be a place for a dedicated camera with detachable lenses, but if a cheap program can improve noise and increase resolution, what will be the point of larger formats? All the best.

[@pentagramyt417]

@@malikknows3510 As a wildlife photorapher what do you think about Olympus 300 mm f4 and darker forest area (I mean, as in forest... there is usually no light)? What ISO could be expected to 1/500 shutterspeed? Honestly for my APS-C sensor, 1/500 and f6.3 is on the 12.800 ISO when it's on "auto" to that range.

[@malikknows3510]

@@pentagramyt417 Thanks for the question, I don't have that lens yet, so I'm not sure. My OM-1 can handhold very long exposures so that helps to bring the ISO down. I rarely need to shoot that high an ISO in my own photography. I do think that 300 mm f4 is a great lens for the money. All the best.

[@pentagramyt417]

@@malikknows3510 Thank you Sir! And could I know at the end, what times could be possible with that lens? Around 1/60? Have a good day though!

[@TheFlyingDogFish]

If anyone ever argues with me about sensor size and noise, I'll send him to this video.

[@lenzflyfishing]

this has been very helpful and I was blown away at the end by seeing that iPhone stack. I have seen my phone do this but I didn't know it was actually stacking. Great video, makes me think as a new "hobbyist".

[@simon_dentremont]

Great to hear!

[@rlgenge]

A great topic, and one that will continue for some time. Using your water and photon analogy, I would add that the degree of 'dirty' water that arrives or 'noisy' photons that arrive at the sensor makes no significant difference to the digital noise seen in the result of the final image as a result of sensor size. It's about the relationship of sensor performance and conversion to digital signals that has the biggest impact. The quantity of atmospheric noise (low light will give you a poorer SNR) arriving at the sensor (assuming equal external conditions of light, lens and camera settings) is equal, and therefore not dependant on the sensor size. Analog-to-digital converters (ADCs) are used in digital cameras to convert the analogue signal from the image sensor into a digital signal that can be stored on a memory card. The type of ADC used in a camera depends on the camera’s sensor and other factors (which you touch upon). For example, some cameras use a single-slope ADC, while others use a dual-slope ADC (better performance) or a successive approximation ADC. The choice of ADC can affect the image quality, especially in low-light conditions. However, the specific ADCs used in different cameras are not typically disclosed by manufacturers, and along with the sensor will have a significant cost impact. Cropping is a form of amplification and it is expected to see more clearly any noise that is present in the final image.

[@simon_dentremont]

Thanks for the comment. I would add though that in your “dirty” water analogy, while the amount of light increases in a linear fashion with larger sensors and longer exposures, the dirty noise only increases at a decreasing rate (square root of the number of photons), leading to a better signal to noise ratio as more light is added.

[@vonchef14]

Thanks for clearing that up, enjoy your work 👍📸

[@richwoodham3296]

Wow! Simon...! Excellent video with lots of detail. Am going to have to watch this again and again to take lots of notes .... LOL.... Thanks for sharing your knowledge and expertise .... 🙂

[@simon_dentremont]

Many thanks!

[@geraldbraun6267]

Great video Simon, I use a micro four thirds camera and you're 100% correct on using a faster lens. I purchased a 2.8 lens, then a 1.4 and finally a .095. The reduction in noise is amazing using these lenses compared to the kit lens.

[@simon_dentremont]

Good stuff!

[@alvaro5162]

Your explanation would be fine if the total amount of energy received by the sensor was related to the noise but that is not the case, what matters is the amount of energy received in each pixel and that depends on the pixel density of the sensor and not on its size. To explain it simply, if we completely cover the surface of one square meter with 12 large glasses and another square meter with 24 small glasses. It is obvious that the total amount of rain will be the same for all small or large glasses, but in each large glass we will have twice as much water as in a small one.

[@simon_dentremont]

True, but it stole the water from the neighboring glass.

[@alvaro5162]

@@simon_dentremont Depth of field also affects image noise because a sensor with greater depth of field allows you to use larger apertures. To obtain the same depth of field with an FF as with a micro 4/3 you will have to use f5,6 instead of f2, those two stops that you gain in the the aperture can be subtract from the ISO.

[@LuisRodolfoCorona]

Extra bonus tip get DXO PURE RAW and let’s the software do some magic for high iso for any camera in raw files 😊

[@umfilmmaker8253]

This is a great topic! I still use a Canon 7D mk1, as well as a 5D mk3. The 7D has been excellent for added reach with the crop sensor for wildlife photography. My Canon 100-400mm f/4.5-5.6 then becomes long enough to reach elk, pronghorns, and bighorn sheep, without breaking the bank. I also love the 5D for lower lighting situations when the wildlife is closer, but also for landscapes to get the fullest effect of my wide aperture lenses. New tech is great, but in 2023, this has still been an excellent pairing on a budget. BUY USED 😊

[@simon_dentremont]

Having one of each really is great.

[@jorgemoro5476]

Fuji 50r and 100s for me. Nothing else matters. Thank you. Printing 70” x 60” easily.

[@michaelschneider9710]

One thing to factor with sensor size is stacked vs. non-stacked sensors. Olympus OM-1 is a stacked micro-four thirds sensor at 20mps, and Panasonic G6II and GH6 are 25 mps, but they are not stacked. I love Olympmpus with 300 f4, and a friend has the 150-300. I also use a Canon R5/6 for a full frame. Since the OM-1 has landed in my gear, with the stacked sensor and new autofocus, and much smaller long lens sizes, I don't grab the R5 and 500 F4 as much as I used to. I am hoping Sigma finally brings their Fovean Sensor II out, it will be very interesting to see what that does.

[@dcd]

Great explanation. I don't think there is any disagreement, larger surface area offers an advantage vs. noise given all other attributes are the same. Ie. same generation of sensor, same size photo-sites, same processing, etc. The problem is when comparing products from different generations which is why a lot of 'switchers' from older FF's are impressed with modern APCs. Sure, they are good vs. older FF sensors but there is always an low light advantage to collecting more photos with larger sensors... One may eventually make the argument with computational photography. If you can scan a small sensor 10x in 1/500, you might get better data than with one scan of a larger sensor at 1/500 for example.

[@davidcrossley7145]

Once again I’m thankful Simon for going over this noise,sensor size.Because I’ve been contemplating purchasing a canon R5 for some time now. Thank you because you’ve helped me make an expensive decision much easier. I’ve decided to stay for a longer period of time with the equipment that I have currently own. So thanks for your incite much appreciated 👍

[@simon_dentremont]

Enjoy!

[@WolferAlpha]

This tip about stacking smartphone photos to get better results is good... I think mine does this automatically when I take a photo in 14 Bit RAW

[@Endureromex]

What a neat explanation, thank you. I recently bought a G lens (full frame) for my 6400 (apsc) camera, just because it had a huge discount. To my surprise the image quality is much better.

[@simon_dentremont]

Great to hear!

[@Warrior_Resisting_Colonialism]

A great advantage to using full frame lenses on crop sensor bodies is, you use the BEST part of the lens (the center).

[@cookshok2014]

Great 👍👍👍 my dear friend 🙏❤️🌺💐

[@pseudophotog]

Excellent explanation - thanks Simon

[@danncorbit3623]

Another important factor (mentioned in your video) in noise and overall sensor quality is the generation of the sensor. For instance, the Olympus OM-D E-M1 Mark II is rated above the Hasselblad H3DII 50 in every single category except portrait (color depth), because even though it is a tiny fraction of the size, it is many generations improved in technology. So buying an old camera with a very large sensor size won't necessarily fix the noise level problem. That same Olympus literally trounces the full frame Canon 5d in every single important measurement. Now, all other things being the same, a larger sensor does produce cleaner pictures. Even so, Topaz denoise can work wonders on a noisy image. Naturally, there are limits, and post processing is not an answer for low light.

[@grattonland]

Nice to see there's experts on the other side of the Bay of Fundy. d'Entremont, you're likely a Par-en-bas. I'm a Lanteigne from la Baie des Chaleurs. Anyway, discovered you this week, watched a dozen of your videos and learned a lot. Thanks for your content.

[@simon_dentremont]

J’y suis! merci!

[@dominiclester3232]

Nicely explained, thanks! Your bound to know, but it was confusing to hear you quote Medium format as being one size. The Fuji/Hasselblad (medium format) sensors are quite a bit smaller than the Phase One sensors.

[@acemanNL]

Thumbs up for Simon! 👍👍

[@nightdonutstudio]

That what m43 fans always say f2.8 is f2.8. Same light hit the sensor as your full frame f2.8 lens. but look at m43 lens size. F2.8 lens are much smaller than full frame lens. Look at depreview comparison. If I lower m43 iso two stops below full frame's iso. I got similar quality image. That kind tells the story. You can acheive great low light with m43, but you need really fast glass to counter it. If full frame is at f2.8 at iso 12800. You need to use f1.4 at iso 3200 for m43 to get simialr quality.

[@obscurelight]

Get your wildlife telephoto lens at f1.4 and see how much you going to pay compared to f2.8 . It defeats the purpose of having compact and cheaper lenses in the m4/3 system compared to full frame.

[@dogpadogpa]

f2.8 is f2.8. The quality changes depending on how you compare. Full frame f2.8 ISO 12800 vs m43 f1.4 ISO 3200 is similar IF the view sizes are the same (same size print and same tech/pixel size). If you enlarge the full frame picture to 4x the size of the m4/3, the ISO noise will look the same (full frame f2.8 ISO 12800 will look like m43 f2.8 ISO 12800). On dpreview try this: Sony a7R V RAW 12800 vs OM-1 RAW 12800. (in theory Sony should be 80 megapixel to be exactly like OM 20 megapixel). Compare = same size print, the Sony looks better. Full 100% = larger full frame view, both ISO 12800 look the same.