Get your Squarespace site free trial with NO CREDIT CARD needed! squarespace.com/maxyuryev Lens used on R5 (Amazon)➡ geni.us/dAPtCC Lens used on A7S III (Amazon)➡ geni.us/qYIi Order an A7s3 ➡ bhpho.to/2CWAHMH Order an R5 ➡ bhpho.to/3fe0sWM

There is pretty clearly an advantage to the Sony, but the shot of the path that seems to have very different lighting between the two cameras. I don't doubt that the Sony has the edge but it's lighting has so many more shadows on the pathway I'm not sure this shot is telling us much.

Lots of people shit talking both cameras when the fact of the matter is, most of the people buying either of these will have no idea how to properly use them to their full potential. There are people on youtube shooting with cinema cameras and their work looks like it was shot on a T3i. The bottom line is, buy the camera that suits your needs and call it a day. Learn to use your camera, and no one will tell the difference as to whether it was shot on an R5 or a7siii. They are both great.

R5 in RAW uses CLOG 2 gamma & smokes the a7siii, I've used both. The R5 is also getting a lighter RAW codec in Nov and CLOG 2 & 3 for all modes. A7siii uses aggressive NR for its lowlight and still has gross green and yellows in skintone highlights. You also can't use it as a decent hybrid, because the MP are too low.

I can't help but feel that Canon's approach is (and has been) getting pleasing results and an easy experience no matter who's hands the camera is in. People always mention "Canon colours!", and I think Canon do always have a Canon-look, whether desired or not. Sony, otoh, has been on a more technically innovative crusade, pushing more functional boundaries for those who want to develop their own look, which may mean not as pleasing defaults shot oob, but gives a broader canvas to work with.

Nice comparisons Max!

From what I notice by first glance, some of the tests are flawed because the shots are at two different times (clouds, whatever) such as the shots at 2:18 .

Is the Standard profile on the Sony plain untouched? Getting a great image more or less straight out of the camera is worth a lot!

These match my findings in the tests that I've done, as well. Since I'm always in Log with either camera, and never the standard profiles,, they are very close for all practical purposes. In the most extreme cases, Sony has a little more dynamic range. But I expose each to protect the highlights and just bring up the shadows in post. I've found the Canon to have finer grain and actually be cleaner through 6400 ISO, so we can basically get equally good results from either. Thanks for sharing. Cheers!

Did anyone notice in the 4:08 the tree shadows on the street are NOT the same !! that means it taken on different times of the day, and they gave the best time for Sony (More tree shadows on the street). I'm a Sony guy but this is kina not fair compression..

You need to change RAW to C-log2 to get the full dynamic range of the sensor. C-Log1 caps at about 11 stops. In Resolve, you have the ability to pick which C-log version you want in the RAW tab (note: clog2 has more DR than clog3). Still, I expect the Sony to have more DR. Regarding colors, the Canon wins, unless you shoot HLG3 with the Sony (with the rec709 color profile). HLG is highly standardize, so Sony can't destroy the colors and get brown/green tints as it does with s-log.

The hard to edit and playback files from the canon are a bigger issue than dynamic range.

I’m really surprised about the color difference. I’m a Sony fan boy and this is the first time I’ve preferred the Sony color over Canon.

I am a professional photo editor that usually edits his own A7-III images. A client with an R5 asked me to do a large number of edits and I notice how the blacks are crushed very rapidly (and not very recoverable in RAW). Worse, the highlights are practically DOA : If they're blown out to white, then even highlight recovery often gives very very minimal results beyond that. It's very disappointing compared to the A7-III in terms of recoverable shadows and highlights.

You need to shoot 8kRAW and switch to CLOG 2 in premiere. (its only possible in RAW right now). Should give you another two stops. Tested the other day. Saved a little more of the highlights. Thanks for doing this video! Good to see the differences in profiles!

8:10 This image reveals the differences nicely because its tonal range is high. The Canon image slightly clips both highlights and shadows equally in a good exposure; the Sony is slightly underexposed, but still avoids clipping in shadows. This reveals good tonal (grayscale) latitude for the Sony that retains more editing options than the Canon that's slightly 'blown-out' in the extremes of image exposure. But then there's *color!* The Canon records the colors with subtlety and fidelity that are much better than the Sony. The Sony has decent blue response, but red & green are too saturated. Post-processing to get colors right can be a nightmare because it involves *two* variables; tone and chroma. Between these two cameras, the Canon has an edge for the photographer who can control lighting and exposure for simpler post-processing success.

How about the fact that there is DCI 4K in R5, but not in A7S III?

The thing is it's not the R5 performing poorly. It's the lack of C-Log3. We already know that C-Log1 cannot utilize the R5's full range so I don't really know why people blow this out of proportion so much. Wait till C-Log3 comes and I would bet it will be a lot closer. I'm sure the Sony will still edge ahead but not by much.

I like how you try really hard to be unbiased. I can see it's difficult 😉

Great video, thanks! I've been playing with the R5 profiles. Canon offers a free editor that lets you tweak them and push shadows/highlights. I had similar results as you using Standard. But after lifting the shadows with a simple S-curve, the results were dramatically improved. I was really surprised at how much detail came back. I also shoot 8K+MP4 (proxy)...and the proxies with the edited profile rival the 8K original.

So why do these images look nothing at all like my R5? ...not even close. Make sure you are shooting in 4K HQ mode not standard as Standard 4k is about half the Quality of HQ mode, there are a few YT vids on the difference between the 2.

Easy math: 13 / 11 = 1,18. Means: The SIII has 18% more dynamic range than the R5. That's what we see here.

The R5 is surprisingly doing well! I still think that SLOG is overrated, I rather shoot with HLG profiles as it's much easier to work with

Sorry, the Canon has clearly worser dynamic range when you see the shadow details on the video with the windows and interior, the wood on the tables looks much much more finer and better, when it comes down to shadow details. Also the Canon post grading looks way too much unnatural, it is the typical Canon 12-13 bits pushed via post processing looks like their photographs. There is clearly 1-2 stops dynamic range advantage on the Sony like all the DxO Mark tests for years. When people are saying different things they lie. The truth is Canon will be behind for years, their image sensor tech is clearly behind. There is a reason why on every technological aspect Sony is quite common used for professional filming especially with their CineAlta film cameras for Hollywood film and Netflix series and film use alongside other companies like Arri and others and not Canon for serious film productions.

Looks like the white balance is a bit off. A lot of Canon footage are tinted towards magenta whereas the Sony towards green. And in the rest of the footage, vice versa.

The classic Canon story plays black, with the eos R5, great review Max, I like how objective you have been here 👍

On paper is one thing but in reality, they seem to be very close. BTW Did you interpret the R5 footage as CLOG 2 in RAW ? I believe that would’ve given you better results in RAW.

I never understand this kind of testing. Like, in the 'real estate scenario' the driveway is blown out. Why? The camera has a histogram. Don't blow out the highlights and compare the final image or the workflow. I get the option to use both at my work (r5 and fx3) and while the workflow is different, I find I get wait better shadow recovery in CLog3. So I just expose for highlights and never worry about it. I'm sure there's a reason people test this way but if one of the guys at my work presented a product with blown highlights and blamed dynamic range, I don't think they'd be working with me for very long

No surprise about the dynamic range from the SONY. On a different note though, if both of your shots were shot handheld, the IS on the R5 is incredible.

Are you using CLOG 1 or CLOG 2 for the R5?

Most of the Canon Footage is from a different time of the day when sunlight is too harsh so anyone can understand what happened in the video

Future firmware upgrade due likely in November for Canon R5 - Canon Cinema RAW light addition - 1080p @ 120fps - C-Log 2 and 3 (this will greatly enhance Dynamic Range) One big issue here that has been commented on is how much Canon wishes to enhance their video quality and specs. Canon is BIG in the Cinema space with the Canon C300 Mark III and the great full-frame Canon C500 Mark II at $16,000 USD. If you want KILLER Dynamic Range then get the C500 Mark II plus a boatload of other cinema features that are amazing. One thing that I don't think anybody denies is Canon color science is at a top tier level. - Bug fixes - Enhancement to temperature control - There are comments on more codecs being worked on One thing that needs to be mentioned, the Canon R5 is FIRST a Photographic (stills) camera and by the way, it also does very good video. Canon marketing got a little ahead of themselves. The R5 is the BEST mirrorless stills camera on the market today. And it can hold up pretty well in the video space. You want dominant high-quality level video then get a Canon C300 Mark III or C500 Mark II.

Hey Max, what I noticed mostly was the colors. Sony had way better colors and blending of said colors sir. This was interesting 🤔 to see. I have always felt Canon had better colors but maybe the times have turned. Thank you for your work and sharing. (:

Can you please do an iso comparison with with the a7iii and a7siii like in the past with the a7sii and a7iii? Nobody out here has made one, or I haven't found it, where they find the highest usable iso. Usable as in usable footage that is

Actually, R5 with 45mp sensor, this much dynamic range is not bad I think.. R5 fought well with A7SIII's 12mp sensor, specially made for video work.. Sony's dynamic range is really good, as always.. Also, I'm impressed by Canon R5, the great all rounder camera.. R5 is one of the best camera for photography ❤️...

By the way the last image you can see canon was struggling with colors when under exposed - the Sony was fine

One thing that really hurt Canon initially was only including C-log and not C-log 3. Newer firmware has added C-log 3 but a lot of time damage was already done in terms of building a bad reputation for dynamic range. It’s still not the best but I find a lot all comes down to camera processing. We measure DR a lot of times in terms of noise. Sony uses a lot of in camera noise reduction to help extend S-log 3 to see more into the shadows and have less noise. A lot of these cameras when comparing raw are much closer than we think. Raw has more noise but the point is no crappy in camera noise reduction. We can use better noise reduction in post. There are different log formats for a reason. C-log 1 is more like S-log 1. Both of which were designed for 8bit lower range sensors. Comparing C-log to S-log 3 isn’t exactly a apples to apples comparison. Slog 3 and C-log 2 are log curves designed for a more than 13 stops of range extending as high as 16 stops potentially. The 10bit point of the curve is much lower for both log formats meaning they dig into the shadows more. They also dip more into the noise floor meaning more noise because of this. Cameras like the A7S 3 use in camera noise reduction to help that look better. It is a cleaner sensor however due to only having 4K resolution. The difference once you compare 8k raw C-log 2 using NeatVideo in post however is a lot closer than people think. This is why the R5C internal video feels a lot cleaner and has a lot more dynamic range than the R5. Both use the exact same sensor and DSP. The main difference is the R5C has an active cooler meaning it can use more sophisticated noise reduction in camera. That’s how it’s able to get almost 1.5 more stops of DR over the R5. R5 raw can do the same thing with good post noise reduction. Just takes more time and effort. Raw video form both cameras are pretty identical. Raw video from a A7S 3 doesn’t look as clean either. So yes some cameras out of the box with internal formats can look better. When comparing similar log formats and raw where we get to use the same amount of noise reduction in post all cameras end up being a lot closer than we realize. Besides this is all splitting hairs anyway. A shot should be exposed properly. Protecting highlights is even an odd notion because they likely are not all that important. For example an open window for an indoor interview where someone insists putting the open window in the shot. 12 stops or 14 stops that window still looks like crap. It’s an unbalanced level of exposure and that window becomes a focal point for the eye instead of the subject speaking. It’s just bad composition. It’s ok to let the background clip because the subject is what ultimately matters. If you want to balance the shot better then light the subject to better match the background. Even photographers use flash or reflectors to help fill in situations like that and create a better balance between the subject, foreground and background. Lighting is the key to cinematography. It’s not how much a sensor can do on its own. A good sensor is important of course but we need to stop over emphasizing details that are not really all that important or not a true apples to apples comparison. Higher dynamic range means better rolloff on either end and more flexibility to fix something. It’s not an indicator of one camera being worse than another. Even for HDR delivery the R5 has more range than we can fit into a HDR 1000 or even HDR 2000 delivery. For rec709 it has way more stops than what are possible to see without looking like a lite log format and super flat. We need to emphasize better exposing and lighting and not if camera A gets 0.2 more stops of range which is likely a pointless stop to have anyway. The way log curves work is at the very bottom of the curve there are very few 10bit values used for those stops. Sometimes only 3-5 10bit samples. This is in comparison to the upper stops that may have over 75 10bit samples. Even if a log format has an extra stop of detail down there it’s likely so insignificant it’s really bad to try to lift it up and not get banding or other artifacts. Lifting up that last stop isn’t going 5o make or break a shot that was badly exposed. In some cases maybe it can save the day. It we are talking really poor details down there and they really should not be lifted very high. A bit cleaner garbage is still garbage. Just like a slightly less blown out window is still a blown out window. We are obsessing over a measurement that actually has very little meaning. C-log and c-log 3 see less shadow detail because the shadow curve tapers off much earlier than S-log 3 does. That’s it. The darks are clipped sooner. Has nothing to do with the sensors themselves. That’s why even with the noise it’s better to use C-log 2 raw on the R5. That’s the more accurate comparison to S-log 3.

Have no idea what you are doing with the R5 RAW, but I saw a huge difference between 8k RAW and 4k HQ when I shoot on my R5. At least in Davinci resolve you can set in the RAW settings clog, clog 2 and clog 3, you can also change the ISO settings afterwards, you can play with the highlights and the shadows like with a RAW Photo and you (at least I) destroy the 4h HQ dynamic range of my r5. I have no sony to direct compare it, but when you write the 8k RAW is just a little better than the 4k HQ, then I guess you are doing something wrong. The difference in the dynamic range is the reason why I try mainly to shoot in 8k RAW and why I have no space on my computer and also why I have to buy more storage... I agree that without RAW the dynamic range on the r5 is not so good, but in RAW I absolutely disagree with your conclusion.

I don’t think you can give a true test of dynamic range by just dropping REC709 luts onto the footage, I’m not sure if that’s what you did but you need to treat each shot as it’s own entity and grade it from the ground up to get the most out of it.

I don't think comparing a7s3 with R5 is fair.. We should instead wait for the lower MP camera by canon to compare it with a7s3

Hey mate , can u please make a comparison video of sony a7siii with bmpcc6k

Sony, protect the highlights and live ever happy 😃. Usually shadow look too dark on the tiny screen, but all the details are there when you go to your computer!

The Sony dynamic range is amazing, however the noise on the piano at 7:03 was awful looking. The blown out segments of the canon was much more tolerable than the sony noise on the piano.

Explain why the details in the shadows are so important? If in real life darkened details in the shadows give a more realistic picture! That is, for filming a video, we have to shoot so that we can capture all the details, and then make a color scheme ourselves similar to reality? But if we write in HLG, how can we restore real colors?

Love the sign... whatever things are pure...

Thanks for doing this video, really helpful in choosing which system to choose. cheers!

The canon codecs are more of an issue, even setting aside DR, overheating etc. Workflow is painful.

Check the skintones at 04:52. Evidently the A7SIII shines here. Her face has much less contrast and looks more natural. This also is a result of the Sony's extended dynamic range. The Dynamic Range on the R5 is worse than the Sony A7s Mark I

Something that I didn't see many people address is the fact that the A7S3 is the first Sony Camera that not only placed the Base ISO lower for Slog3 (at 640), but it also allows you to go even lower, to an extended rage. Now they say that going into that range costs a little bit of Dynamic Range. But I want to know how much DR is lost. Is it noticeable? Is it useful if you just want that extra Stop lower, because your ND only starts from 2 stops and you just need 1? Or is it better to put on the 2 stop ND and go ISO 800 to achieve the same result?

i have a question, if you guys have opportunity to meet a guy like steven spielberg, or christopher nolan, would you ask some question like, "are you concern about dynamic range when shooting E.T ?" or "what log do you used when shooting memento?" i mean, when i'm shooting horror movie, i dont want everything in the scene is lit ...

What I'm more happy to see in this review is how close the cameras look in terms of color. People have always talked down on somy color science and this is a good reminder that its been much improved in many ways. The a7siii is an amazing camera , bar none.

I was wondering what your exposing strategies are? Are you exposing for the same anchor point like a grey card or skin tone for each camera? If not, the results mean pretty much nothing in most cases as you can always adjust them to look roughly the same unless absolutely clipped. I’d be curious to see both cameras exposing to the same back lit face according to official guideline(skin peaks at around 60ire, or middle grey at 41ire for slog3 which I’m familiar with) , then corrected with official luts and nothing more. That would be the best simulation to a real world use. Without being a fan boy I really think we should be seeing a massive real world difference considering clog is artificially clipped at 12(?) stops.

Great comparison. Thanks for sharing. Cheers from LA.

This is super interesting for someone trying to explain to his wife about dynamic range and how 2 of the best current cameras compare. Much appreciated also are your honesty and impartial statements. We (my wife and I) noted that in the interior footage (e.g. 8:06), while the highlights are blown-up with the canon and slightly better with the sony, the wood on the left part of the piano is slightly better with the canon. The tree on the right hand side looks slightly more natural (more contrasty and greener) with the canon. All in all, without such side by side comparison, it would be hard to tell that the Sony is better than the Canon or vice versa. What counts, and you show it superbly, is who is shooting and how well is the camera mastered.

Remember to get 10-bit you have to go log. So, standard profiles on the Canon are going to be the 8-bit files.

CineD has the Sony at 12.4 and the Canon at 10.8. A delta of 1.6 stops. Their is testing is quite a bit more scientific. The Sony has some other issues too. If you want to use the low light you really need to shoot to external RAW. And how does the Sony work with respect to your workflow to delivering to commercial customers? It is certainly good at super low light. But lots of professionals are using other full frame cameras and are getting very deliverable content. If you need 150,000 iso then it's a one trick pony. If you need something that's more realistic say to 12,500 there are lots of other options out there. And if you do heavy grading, you have to be careful with which codec's you're picking on Sony. If you're in the Sony ecosystem and this checks off boxes you need it makes sense. If you're not, it's not so cut and dried.

Is the Sony A7S III Dynamic Range as same as thr new 7iv?

Are you going to do a low light comparison between the A7SIII and the A7III?

I must say it looks like Sony finally has matched or even surpassed canon in the color department besides the advantage in dynamic range. Plus no overheating issues. Don't see any reason to go with Canon.

Sold my A7s3 for R5. Never look back...

There is Half or a third stop difference if that between the 2 cameras, there is no way a 2 stop difference

I wonder how the comparison would be if you shot in _Prolost Flat_ instead of Canon's standard profile. Canon always has way too much contrast and saturation.

well we have to wait for canon to add clog 3 so it is still an one sided test

You didn't dial in any exposure on the R5.

Cannon has a cleaner image. Colours better on Sony. Look at the grass on the rear RHS of the track. That being said it could be the f value.

Thanks. This is THE video comparison I needed!

Canon may be lagging in some video aspects but it's pretty all rounder it can do both photo and video but I can't even think of photo side to sony a7siii... Keep in mind sony is dedicated video camera so it should have that edge

There's no fix for Canon's standard profiles, no tweaks to bring it up to Sony, just clog or crap?

Hopefully new Clog3 from canon will close the gap a bit. The real headache is working with the files. Insane. Maybe ARM based MacBooks will change thst.

You are the MAX bro !!! . Nobody Does what you do !!

dont ever say "better colors" that statement is quite ignorant. Better color balance, white balance, more accurate skin tones, yes that you ca say. Better colors is not a thing, like PB always says, you can make anything look like anything if you put some effort into it. We are soon past "canon having better colors than sony" stage at this point.

Will always go with canon because even though sony has better dynamic range, canon r5 is a proper hybrid camera which unfortunately sony is not

Will you get your hands on a Nikon Z 6II?

Canon standard profiles were always bad. Regarding your test they don't seem to have matched diaphragm in the woods, the DOF seems different, but that shouldn't impact anyway. And definitely A7S3 has an edge there, probably due to the BSI.

from last scene that is only like half stop worse in highlight to me.

Will c log 3 make a better difference?

Great review. It still comes down to the fact that we are comparing a photo centric with a video centric camera. Canon, for a photo centric camera is doing quite well if that difference is so small as shown by your examples. Thanks for the review. Well done

Why not also try RAW with the a7siii?

Amazing comparaison. Thanks!

Sony A7S III or Blackmagic pocket 6K??

You cannot compare them using luts, that driveway woudln't clip if you did it manually.

In the beginning, you were comparing dynamic range using standard profiles? Why? That's why I had to stop watching the rest.... Not sure what the point of that was?

DOES A7S III CAN SHOOT DCI 4K EXTERNALLY WITH NINJA V ????

How to handle c-log in Davinci Resolve if I want to run Color Managed (or via ACES)? I can't find a profile/IDT for the R6/R5 for c-log. Would prefer not to use destructive LUTs.

Great review!

Love your video. I own an R5 and i hate canon's dynamic range. But I hate sony color.

Can you make an ibis handheld comparision for travelling shots without a gimbal in both cameras? Thanks!

I still think that the R5 is a better hybrid camera. That dynamic range difference is not really going to matter that much to me. I’m not sure that the filmmakers that I know and work with are using these for that anyway. I see them using cinema cameras. Overall if I didn’t have all of that canon glass,k and if I didn’t need to take pictures as well, I would probably get the A7S3.

Where we can find this lut v2 for canon please. Like always you nailed it 👌🏽 thank you

Hello, I found YOU man. AWESOME videos. EXPLAIN a lot! If you are wondering who I am, then I am related to Aunt Liliya Bogdanov. TIM here!

Ooooh!! The R5 takes another huge blow!!! Totally BOMBED!!!

That Cripple hammer is running rampant all over the R5. What a shame Canon LOL

Of you do a color space transform in resolve to the Sony log footage, you’ll get stellar results. A LUT isn’t going to get you all the way there in quite the same way.

I wonder if Clog3 would improve this if Canon releases this as they’ve said they will.

Dude you are over exposing Canon R5C. By the way, you can retain Highlights better in R5C and Shadows in A7S3/FX3. Do your tests properly first.

8:35 Save some for the rest of us.

I believe if people are complaining about how cameras perform in 2020 then they should switch jobs/hobbies

Now why do so many of those shots look so familiar to me? 🤔 good video man

Definitely the Sony A7III has better dynamic range than the R5 for the simple fact that we are dealing with larger pixels at 24MP vs the 45MP of the R5. I own both and the R5 have been limited to stills which it outperform the A7III but when it comes to usable video even at 8bit the A7III has better DR than the Canon R5 with it's 10 bit. Great comparison Max.

The test Gerald did was pure sensor test... no LUT was applied. The sensor can’t resolve more than 11 stops