RVB
Member
I'm not sure how much value there is to be found in DXO's testing but this may be interesting to S/S2 owners.. Leica S sensor review: Consummate performer? - DxOMark
The GetDPI Photography Forum
Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!
DXO publish their Leica S test..
- Thread starter RVB
- Start date
Kolor-Pikker
Member
I am now disillusioned, and all those great S/S2 shots that I've seen to date are actually bad both objectively and subjectively, all thanks to a vague test that I hardly understand that I've read on the internet, since the other cameras clearly have bigger numbers. 
DxO measurements can be useful and provide a means of objective comparisions for a number of criteria but like those measurements in audio, they often don't reveal "all" what the eyes (or in the case of audio...ears) experience in terms of visual or aural stimulation. Sometimes a image somewhat limited in dynamic range can often project more impact than a flatter wider dynamic range image, even after appropriate post processing.
There are so many underlying factors that go into a visual image as well as a audio recording, so I take these sort of measurements with a healthy and measured does of respect but don't consider them the end all and be all in my decision making process in what to purchase.
Dave (D&A)
There are so many underlying factors that go into a visual image as well as a audio recording, so I take these sort of measurements with a healthy and measured does of respect but don't consider them the end all and be all in my decision making process in what to purchase.
Dave (D&A)
I had the Pentax 645D (which has a similar sensor as the Leica S) and compared to my Canon 5D mark 3 I found that it had about a stop more dynamic range in the shadows but less in the highlights when exposed normally to rated iso (both cameras) so dynamic range is very close but I would give the slight edge to the Pentax/Leica. Color depth is about the same but color differentiation seems better on the Kodak sensor. Resolution is about the same on the Kodak sensor as the D800E.
So my experience kind of confirms what DxO concludes.
The big advantage of the Leica are the Lenses. They are all superb from what I've been able to gather. The body itself is also very well made and feels like a dream in hand. The sensor is a bit dated but still competitive.
Compared to the Phase One IQ160 which I own, the Phase is in another WORLD of Image Quality. Dynamic Range, Color, Resolution, etc. All superb.
Keep in mind that when comparing the dynamic range figures that DXO publishes on the PhaseOne backs I checked and they are NOT done at base iso on the PhaseOne backs! They are done at the iso 100 setting which is one stop above base iso. (maybe their test software/hardware has iso 100 as a lower limit for testing dynamic range) The Phase backs have the most dynamic range at base iso.
RVB
Member
I love the S as a system,great glass,great viewfinder and nice ergonomic's and the leaf shutters are a real bonus,I haven't used a 5D mk3 but had a 1DX and the Canon sensor wasn't very good at handling shadows and didn't like being pushed in post..
Considering the Kodak sensor is getting on a bit It did pretty well,I still love the colour from it but I'm really interested in seeing what Leica offer at photokina this year.
But agree that the Dalsa sensors are better,I would be more than happy if they appeared in the next S,Although CMOSIS is the most likely.
Rob
ondebanks
Member
From the DxO review: "Given the Leica’s slight advantage in pixel pitch, the 41-Mpix ‘cropped 645’ (33 x 43mm) size CCD in the Pentax performs very well indeed."
You can tell that someone hasn't had their Weetabix when they make two factual errors in the same sentence! The Leica and Pentax have the exact same 6.0 micron pixel pitch, and the Pentax is 40 MP not 41.
Ray
You can tell that someone hasn't had their Weetabix when they make two factual errors in the same sentence! The Leica and Pentax have the exact same 6.0 micron pixel pitch, and the Pentax is 40 MP not 41.
Ray
ondebanks
Member
The only thing that I find surprising or disappointing about the Leica S test is its much lower "Sports ISO" than the Pentax 645D, i.e. poorer low light performance. It is supposed to have the same Kodak 6 micron, microlensed technology as the Pentax, so we may assume that it is equally receptive to light signal.
The implication therefore is that the Leica has higher readout noise than the Pentax...its somewhat lower dynamic range score would also imply this.
That is really not good enough for Leica engineering.
As for their comparison with the Sony CMOS cameras (A7R & D800E), what the Leica gives up in the DxO scores up is entirely due to the lower base ISO readout noise of Sony's CMOS.
Ray
The implication therefore is that the Leica has higher readout noise than the Pentax...its somewhat lower dynamic range score would also imply this.
That is really not good enough for Leica engineering.
As for their comparison with the Sony CMOS cameras (A7R & D800E), what the Leica gives up in the DxO scores up is entirely due to the lower base ISO readout noise of Sony's CMOS.
Ray
dougpeterson
Workshop Member
Any attempt to determine the quality of a photographic camera with only numbers is doomed to tell a very small part of the relevant story. And it may not even tell that part accurately.
And it may not even tell that part very well. The things that DxO measures and the ways they measure them are not especially great for many photographic applications. They ignore aesthetics of noise/grain (as important or more important than the absolute engineering quantity), the contribution of the rest of the workflow including raw processing (e.g. not taking advantage of the dark frame technology Phase One uses for shadow quality enhancement), and equalize across different resolutions in ways that are counter to answering the questions most people here would be asking. Not to mention they measure color in a way relevant only to a very very small number of photographers. I've written pages about it before and don't care to repeat that experience.
But suffice it to say if you are considering an A7R, D800, Pentax 645D, IQ180, or Leica S (the cameras listed in this dXo article) you will be FAR better served by testing those cameras, or by looking at raw files, than you will be by reading that article.
I say this despite the brand I sell (Phase) having bested the camera under test (Leica S).
And it may not even tell that part very well. The things that DxO measures and the ways they measure them are not especially great for many photographic applications. They ignore aesthetics of noise/grain (as important or more important than the absolute engineering quantity), the contribution of the rest of the workflow including raw processing (e.g. not taking advantage of the dark frame technology Phase One uses for shadow quality enhancement), and equalize across different resolutions in ways that are counter to answering the questions most people here would be asking. Not to mention they measure color in a way relevant only to a very very small number of photographers. I've written pages about it before and don't care to repeat that experience.
But suffice it to say if you are considering an A7R, D800, Pentax 645D, IQ180, or Leica S (the cameras listed in this dXo article) you will be FAR better served by testing those cameras, or by looking at raw files, than you will be by reading that article.
I say this despite the brand I sell (Phase) having bested the camera under test (Leica S).
I am neither surprised nor disappointed (as a S user).
I never doubted that the sensor of an IQ 180 would be better than that of the S.
Also the S sensor is smaller than the other MF-backs, another disadvantage in regards of pure IQ from the sensor.
For me the S is a "in between system", making a small compromise compared to the big backs in regards of the sensor, probably compensating some part of this with the great S-lenses.
But on the other side you get a relatively fast, weatherproof, "compact" system.
I never doubted that the sensor of an IQ 180 would be better than that of the S.
Also the S sensor is smaller than the other MF-backs, another disadvantage in regards of pure IQ from the sensor.
For me the S is a "in between system", making a small compromise compared to the big backs in regards of the sensor, probably compensating some part of this with the great S-lenses.
But on the other side you get a relatively fast, weatherproof, "compact" system.
Ditto.
My (now sold) H4D/60 produced better files than my S2P, as it should have being 645 60 meg Dalsa. However, "better" is tough to define in absolute terms when evaluating any creative endeavor. The science guys need to set some degree of measure I guess, but it is merely a guide to match potential needs against possible performance … the rest is up to us as creative people.
I've found the gestalt of the S to be its unique strength … the combination of the S lenses with their specific signature/rendering, how the sensor delivers against my aesthetic expectations, and the simplicity of use of the whole system, makes it ideal for what I look for.
Perfect example of this is how the flawed Leica M9 produced the sort of look and feel I loved, and the much touted M240 did not. I found the S2P a perfect aesthetic companion to the M9.
In other words, it all looked good on paper, until the paper was photographic
- Marc
Last edited:
Mark Gowin
Member
The S sensor is ancient in digital camera years. For example, I have owned my S2 for 4 years now. I know the DxO numbers are not "competitive" in today's market, but I know better when it comes to real world use.
GrahamWelland
Subscriber & Workshop Member
I guess this is all interesting for the measurebators ... the real proof with all these systems is what folks produce using them. I can't say that I've ever seen a Leica S image that was bad because of the sensor test performance.
Evidently Leica S shooters should dump their over-priced and underperforming systems and get an A7r or D800e apparently. :facesmack:
Evidently Leica S shooters should dump their over-priced and underperforming systems and get an A7r or D800e apparently. :facesmack:
ondebanks
Member
Now Graham, you know better than this "measurebator" name calling.
System characterisation/quantification in absolute (and more usefully, relative) terms has a valuable place in what we do.
I have seen Leica S high-ISO images which, while looking quite good, would have been better if they had the much lower readout noise of an A7r or D800e. And indeed that goes for all CCD MFD units, my own included.
I have also seen Leica S long-exposure images which, while looking quite good, would also have been better if they could have been exposed for longer than the hard firmware limit of 125 seconds, imposed by dark noise which is again higher than a modern CMOS sensor. And indeed that is a problem for most CCD MFD units, my own included.
The limitations I describe does not mean that anyone should "dump" their Leicas.
But it does mean that there is substantial room for improvement of MFD sensors, for quite a range of shooting applications.
An awful lot of us seek the best of both worlds: the resolution, colour, and highlight quality of our MFD systems combined with the low noise (high ISO, shadow detail, and very long exposures without darkframes) of 35mm CMOS. It's not there yet - the IQ250 disappoints somewhat, even ignoring cost -, but it's getting closer.
Ray
RAY, I think we have to watch what we wish for.
In the headlong rush to become a "Jack of all Trades", many of these super cameras have become a master of none.
For every advancement there is some sort of penalty … and technology hasn't solved that yet. Unfortunately, no one can compete with pipe dreams of what MAY happen … all we can do is deal with what's here now so we can take photos … now.
As Dave mentioned earlier, improved DR brought with it flatter tonal range that often cannot match the aesthetic impact of a "lesser" choice … even with astute post work. Plus, it's not like the "lesser" solution is all that bad, just a possible issue in a small set of circumstances … which can be thought of as a "tail wagging the dog" syndrome … lose the impact for 95% of the shots to get a bit better DR for the remaining 5% that would benefit from it. :facesmack:
Same for high ISO, while noise has improved a lot in a number of different cameras, none has solved the color fidelity going to hell in a hand basket.:thumbdown:
The manufactures have to love all this … it is constant stream of meeting needs that people think they want, so we get to decry last year's advancements as outdated, and plunk down skids of cash to get this year's solutions. Meanwhile everyone is scratching their noggins wondering why their work isn't any better than last year.
BTW, I have a A7R, use it with M and FE lenses and love it … it is no Leica S2/S and never will be.
- Marc
Guy Mancuso
Administrator, Instructor
Actually this is correct the Leica has a older sensor like the P45 and that series of backs that use the Kodak sensors. I know they are different but there look is not . Kodak does have slightly less DR. Than Dalsa and the newer CMOS chips and the S is one of those that have less but also Kodak sensors have more bite or crunch to them . Which gives it the punch that many people like. Nothing wrong with this and actually a feature. So yea the data maybe looking bad but the reality is given the glass as well a very well balanced image with a certain look to them that many people enjoy. If you want to relate it to something its real easy think Kodachrome which has a lot of punch to the files and color compared to negative film. It's pretty much the call on it. Some folks love the results. So I totally get why many owners love it. Besides all that the glass is outstanding.
My analogy you can read the box but how does it taste is what counts.