NMOS v CCD - my experience.

[Ted C.]

bringing up spelling errors often appears to be the last line of defense when one has had his position weakened on the forums. That's what the laughter was about, a poke at myself...

I'm done with this post/thread, have fun, come on back, ya hear!

Just having some fun jay bo, I'm sure you will hang around to help
us identify our new cameras' sensor issues. Since you are
obviously new to our forum, I'm just giving you a little dose of
what will surely come if you choose to be our new resident canon
fanboy/gearhead/troll.

If that's not your intentions, give us something to chew on man.
So far, you've come across like a lightweight gearhead with a new
toy. Believe it or not, plenty of this type have come and gone
through here since the release of the E-1.

Credibility is earned, not given lightly by most on this forum and
others here at DPR. That's why you don't see anyone here
clamouring for Louis to prove his point. He's already done that
countless times, even at the expense of Olympus and its users
(though I haven't always agreed with how he reached his
conclusions)....

Come back soon, ya hear!

I will continue to look forward to your posts and anxiously await
the plethora of knowledge you will no doubt continue to provide us
(and the other forums) along with more images of towels, egg
crates, bottles, etc.

After carefully examining your test shots, there is no question
about the superior noise handling characteristics (and other exotic
sensor thingies) of your camera!

Have a good day jay bo....

But Ted, if you can show me how my skill as a photographer impacts
on the validity of my assertions regarding sensor dynamics, then
I'll promptly zip up my 50 best 5 star images and upload them to
pbase.

I await your cogent reply.

I await his test results.

Seriously though Louis, it should be easy for you to demonstrate
this effect for us. Post identical subjects taken under identical
conditions, and we'll all see what you are seeing, (presumably.)

Here's a little test I performed the other day. Where's the sheen
here?

http://forums.dpreview.com/forums/read.asp?forum=1020&message=21904280

even if you turn the ISO up until there is some noise, it STILL has
that slight sheen. So, no, it isn't lack of noise.

However, it isn't a horrible thing - some people may even like the
look. I'm happy with it, but I'd rather do without it.

As for colour - I edit the bejasus out of colour. I'm talking
about something else, though what it is exactly I'm not sure (if it
was obvious what it was it could be engineered in or out).
--
http://www.flickr.com/photos/acam

--



'How one responds to failure, not success, could be the better measure of character.'

 

[Rob Paul]

I agree that they should have tried to move a bit faster but, well,
if the 20D is any indication of things it's that you never ever
rush new product to the market when you are looking to make a
splash while keeping your technology RELIABLE. The E-1 didn't need
no 15 generations of firmware upgrades. And thank goodness.

Just curious about your comment - I've been using a 20D for well over a year now and have not had issues with firmware upgrades - is this something my camera hasn't told me? What is the "15 generations of firmware upgrades" all about? I still use the firmware the camera came with.

Do you use a 20D too?

 

[Jay Turberville]

From reading the past several few weeks of this forum, I can see
that subtle, (or not so subtle) ad hominem attacks are quite the
norm on this forum, so your post isn't very surprising.

OK. Prove it. The part about ad hominem attacks being the norm that is.

You can find ad hominem attacks in just about any forum. That's no surprise. Just as its no surprise that people often over-generalize.

This is actually one of the better behaved groups I've been involved with.

--
Jay Turberville
http://www.jayandwanda.com

 

[Rob Paul]

I'm done with this post/thread, have fun, come on back, ya hear!

'How one responds to failure, not success, could be the better
measure of character.'

I was attracted to this thread because of the heading - "NMOS v CCD - my experience."

On reading many of the postings it seems to me that much of it is defensive in character as are your responses to jaybo who made a reasonable and informative post.
Why is that?

I have a number of cameras made by different manufacturers and read different forums but this is the first opportunity to read postings at Olympus SLR talk. To be fair, I do also have several Canon cameras, one of which is a 20D. It's fine to be enthusiastic about the cameras you use but reacting inappropriately and defensively about the brand as well as denigrating another brand can be a put off for those interested in learning about the benefits to owning an Olympus product.
I will explore the product line more, however.

I like the sentiment stated in your signature...

 

[Olyinaz]

This is generally a pretty well behaved forum, but we've been plagued by a series of troll-like attacks of late and it could be that people are a bit short fused as a result. I make no insinuations about anyone in this thread, just saying...

It doesn't help that Oly users have suffered with a brand that's really easy to like due to uniqueness and optical excellence but boggles the mind due to marketing buffoonery. Frankly, I think this can tend to make us a bit cranky as we wait for our system of choice to finally deliver things like image stabilization and DSLRs with more than 8MP that other brand users have enjoyed for years now, but maybe I'm projecting there.

At any rate, great cameras, only live view, best dust-buster going, optics that are good as any out there - we've got that and some great photographers as well. Stand by for incoming great product from Olympus on top of it.

Best,
Oly

 

[Rexian]

I totally agree agree with you Olyinaz. I never understand this defensive behavior from Oly users, it may be expected from Olympus, the company.

Noise is BAD. How many of you take pictures at ISO 800 under bright light just to have noise ? The grain in the film looks good in some cases, and same goes for noise in digital. It can give an artistic touch, but it should be done in PP where one has total control. Removing noise is a real pain and you risk the sharpness and color.

Canon, Nikon, Oly, no one really had much control over the grains of high ISO in film days and it's probably 1 of the reasons why film is going to be history soon. Oly cameras have many good features, I consider LV as a very important one.

But it's still not the only selling point and that's probably one of the main reasons why Oly doesn't have a good share in the DSLR market. Other than LV, what else do they have that on one has? Size of E-400 is very attractive but it's hardly an option for most of us here :-(

The competition has very clean high ISO (C & N, up to 1600) or in body IS (Sony, Pentax). Oly lacks both these (along with slow AF & horrible VF in E-500). These 2 features have really made digital superior to film and if Oly could have 1 of these (IS is probably coming), they'll have a killer combination for us who don't need a 20MP camera.

Good photographers are good photographers, no matter what gear they use and they can be defensive I understand. But Canon high-end sensors are technology wonders and what's wrong in admitting it. I seriously couldn't see any sheen in those photos posted by Jay Bo and ISO 3200 probably blows most Oly's at ISO 800.

If Oly doesn't announce an E-430 (or E-410 or E-510) with IS & LV soon, they'll be losing more customers :-(

IMO all noise isn't always bad.

Just my personal opinion, but I disagree. One can always add noise
but taking it away is troublesome, as your following statement
alludes to:

Too much noise reduction begins to look too obvious, just like too much sharpening.

Of course. I'd rather have clean signal from the sensor, and on
that level I thinks it's pretty obvious that Olympus DSLRs are the
worst of the lot commonly used brands. Oh well, the other
strengths of the system work for me and the future looks bright so
I'm perfectly happy to stick with Olympus for now.

Regards,
Oly

 

[Louis_Dobson]

His post is about another camera, as I'm just about to point out to him.

I wrote the original post and started the thread, and what I was discussing was the E330, not the Canon 5D.

Does the 330 have a Canon like look (because both sensors have MOS in the name)? No, if there IS a Canon look, the 330 does not have it. Subjectively it seems to out perform the 8MP CCD sensors used in my E500 without being different in nature.

Jay Bo is dragging the thread of into a discussion of the 5D. I actually know the 5D quite well, so I CAN have that discussion, but not here.

--
http://www.flickr.com/photos/acam

 

[Louis_Dobson]

Let's clarify, this is an Oly forum, and this is a thread about the E330.

Many people feel they can see a Canon look. Many people suspect that is the sensor technology. The Canon has a MOS sensor (CMOS). The E330 has a MOS sensor (NMOS). The other Olys have a CCD sensor.

Therefore people may reasonably wonder if the E330 will have the Canon look, rather than the Oly look.

This thread is designed to inform them that, based on extensive use of the E500 (8MP CCD Kodak), extensive PPing of E1 images (5MP Kodak CCD), and two weeks heavy use of the E330 (7.5MP NMOS Panasonic) that the E330 does not show any Canon like behaviour.

The E330 gives typical Oly results, but subjectively a little better than the E500 and up there with the E1.

Now YOUR concerns seem to be bit different - you seem very keen to defend the honour of the Canon 5D. Well, by coincidence, I know the camera well, but here is not the place to discuss it in detail.

You've asked for samples - samples of all the cameras at 100% are available in my galleries as I told you. You've asked me to stand around with all three cameras taking identical scenes. Excuse me while I put on a beret and light a Gauloise - I'm an ARTIST sweetheart, you can do your own silly tests. Actually using the blated thing to take pictures, I know what I find.
--
http://www.flickr.com/photos/acam

 

[jay bo]

You are just not responding to my questions.

I'll say it again.

It would seem that your hypothesis that the factor involved here, the cmos sensor, is incorrect. In fact, it follows from your very experience with the E330, a camera, I should add, I have used as well.

I am trying to point out to you that the factor that you attribute to cmos sensors could in fact be anything in the imaging pipeline. I made specific mention of Canon lenses being biased for high contrast as being one likely culprit. I am not defending any "honor" here Louis, I simply in doubt as to the validity of one of the basic points in your thread.

At the same time, I asked you to point out this "sheen" to which you refer. Show us the sheen. Point it out to us. If you don't want your assertions to be examined with a very keen eye, then I guess that is okay. But you should realize realize that this IS one of your key points.

Let's clarify, this is an Oly forum, and this is a thread about the
E330.

Many people feel they can see a Canon look. Many people suspect
that is the sensor technology. The Canon has a MOS sensor (CMOS).
The E330 has a MOS sensor (NMOS). The other Olys have a CCD sensor.

Therefore people may reasonably wonder if the E330 will have the
Canon look, rather than the Oly look.

This thread is designed to inform them that, based on extensive use
of the E500 (8MP CCD Kodak), extensive PPing of E1 images (5MP
Kodak CCD), and two weeks heavy use of the E330 (7.5MP NMOS
Panasonic) that the E330 does not show any Canon like behaviour.

The E330 gives typical Oly results, but subjectively a little
better than the E500 and up there with the E1.

Now YOUR concerns seem to be bit different - you seem very keen to
defend the honour of the Canon 5D. Well, by coincidence, I know
the camera well, but here is not the place to discuss it in detail.

You've asked for samples - samples of all the cameras at 100% are
available in my galleries as I told you. You've asked me to stand
around with all three cameras taking identical scenes. Excuse me
while I put on a beret and light a Gauloise - I'm an ARTIST
sweetheart, you can do your own silly tests. Actually using the
blated thing to take pictures, I know what I find.
--
http://www.flickr.com/photos/acam

 

[Louis_Dobson]

refering to "my hypothesis".

I don't have a hypothesis, so you are either not reading or not grasping.

What I'm saying is that E330 is not producing shots with a "Canon" look to them.

There are many hypotheses as to where the Canon look comes from, and one of the most common is that is a feature the MOS technology, which the E330 shares. Ergo, one might wonder if they would share the "look." They don't.

I don't really want to get dragged onto your agenda, which is not relevant in this forum, but I'll say this much:

Hypothesis 1 - there IS no look. I disagree, I've handled the RAW files, there IS a look.

Hypothesis 2 - it is the ultra cleanliness of the files. I disagree, even at higher ISOs, where there is noise, the effect remains.

Hypothesis 3 - It is just a colour difference. I disagree, I process colour heavily anyway.

Hypothesis 4 - It is a function of CMOS directly. Possible, I don't know.

Hypothesis 5 - It is a function of the individual pixel cleaning in Digic 2. I feel instinctively this is the most likely explanation, but there is no means of testing it.

Whatever, it makes no difference to THIS thread. Comparing the E330 output to OTHER OLY CAMERAS, the IQ is subjectively as good if not better.

Possibly it doesn't measure as well, and I've not tried photographing a wedding and looking at the detail in the bride's white dress, but overall I think it produces attractive, detailed, punchy photos with minimal PP.

Any comments only apply to RAW, I've not tried and have no interest in the JPG engine.

--
http://www.flickr.com/photos/acam

 

[thebard37]

If you're not laughing, then lighten up!

--



--
Zach Bellino
'I prefer my lo-mein of the veggie variety.'
--ZJB
'There are three kinds of lies: lies, damned lies, and statistics.'
Benjamin Disraeli (1804-1881)

 

[Louis_Dobson]

the system is MUCH smaller and lighter. Not the cameras, but the overall bag.

There is a disadvantage, higher noise at high ISOs, which can easily be removed of course, but there is some detail loss.

Now for someone who does mostly low light work and barely moves the cameras from the car, Oly is a poor choice.

Me, I carry cameras for an average of four hours a day, and spend maybe 1 minute a week running NR.

So for me weight 1,680 times more important than low noise, and I'd not touch anything else with a bargepole as my primary system.

I suspect themajority of buyers, if they were smart, but most people are, let's face it, pretty thick, and can't use a camera anyway, so they'll not be buying Oly any time soon.
--
http://www.flickr.com/photos/acam

 

[vlatko]

Here's a little test I performed the other day.

Just as a point of interest...

I don't think that a shot taken from a metre, metre and a half is really good to test the detail-capturing abilities of any camera. It would be a really poor camera that couldn't capture the detail on the rug and the carton at that distance. Given a proper lens, of course. A coke-bottle lens would smear everything.

And saying that a 12 MP camera can capture more detail than an 8 MP one is like saying... that Mazda RX-8 can outperform a Honda Civic.

D'oh! Of course it can!

But show me a comparison between a Mazda 3 and a Civic and then we can really talk.

In other words, you introduced so many variables into the comparison that the discussion becomes meaningless.

 

[rovingtim]

Hypothesis 5 - It is a function of the individual pixel cleaning in
Digic 2. I feel instinctively this is the most likely explanation,
but there is no means of testing it.

I think this is a likely hypothesis, but I would point the finger at the on-board noise reduction. The problem is it would be impossible for any man or machine to look at the colour of a single pixel in isolation and determine whether the read-out is noise or not. The best one can do is to look for the most likely scenario for noise (ie lots of contrast and quick colour transition compared to neighbouring pixels). There is no way that some real-world randomness is not also caught up in this noise reduction process. The questions is how good is the balance between the cleanliness and the permitted chaotic expression?

I'm going to put this to a test soon, by looking in detail the files between an E1 and 5D. Just waiting for a time to go shooting with a mate of mine.

 

[thebard37]

. . .you are doing this . . .

on the E500 unless I am playing heavily in PP.

It is so true, and it happens with any camera. If you blast the contrast and saturation, then sharpen it up, you will find grain in any digital camera.

Even cameras of the 5D ilk . . .most of the "noise" people think they are seeing are interpolation artifacts, uneven color in the sky etc, that is accentuated by going over the top in PP.

PP work is always compromise . . . if you want super-duper contrast and hyper-real sharpness in all your shots, be prepared to lose DR and increase noise and artifacting.

If they are exposed properly, and PPed well, even ISO 1000 shots can come out pretty clean from cropped sensors . . .you just can't go bonkers with the salt in those recipes.

--



--
Zach Bellino
'I prefer my lo-mein of the veggie variety.'
--ZJB
'There are three kinds of lies: lies, damned lies, and statistics.'
Benjamin Disraeli (1804-1881)

 

[Leon Wittwer]

If people don't agree with you, they are "thick." Sounds a bit condescending if not a bit arrogant. It is best to respect the views of others if you expect some respect in return.

I suspect themajority of buyers, if they were smart, but most
people are, let's face it, pretty thick, and can't use a camera
anyway, so they'll not be buying Oly any time soon.
--

--
Leon
http://homepage.mac.com/leonwittwer/landscapes.htm

 

[Louis_Dobson]

Other people have different priorities and use cameras in different ways.

I very nearly bought into 35mmFF a few months ago because I though I might go shooting stock as a retirement job. With that different set of priorities, I'd have wanted different equipment. If my usage changes, so will my chosen cameras.

As to the more philosophical point, I AM arrogant. 25 years of being paid to advise people on technology has taught me that 50% of people refuse to think rationally, indeed hold rationality in contempt, of the remaining 50%, 50% are too poorly educated to grasp the issues, and of remainder, 50% are simply too thick to have anything explained to them. I have a rather jaundiced view of the human race...

So everybody is entitled to their opinion, but I don't have to "respect" it - in fact, if it irrational, ill thought and dim, I'll sneer openly at it. And if they feel the same way about me, fine, I don't give a stuff.

--
http://www.flickr.com/photos/acam

 

[Scott Whittemore]

I've long got the feeling that CCD was nicer than CMOS, and by
extension NMOS, but not by much.

Bearing this in mind, I expected the E330 to produce marginally
less attractive RAWs than the E500.

It doesn't. After three weeks of use, I'm confident the E330
produces better images than the E500 (yet another reason to read
the E330 review on this site for a good laugh).

So, score one for NMOS.

By the way, this isn't about high ISO noise. I've no idea (nor do
I care) which camera has better or worse high ISO noise.
--
http://www.flickr.com/photos/acam

I agree, Louis. I've said since the E-330's launch that I thought it was the best Oly DSLR since the E-1 and I still believe that. I've all but retired my E-300s & E-500. I still prefer the E-1 but the E-330 comes in handy for certain shooting situations and produces excellent image quality. The E-500 and E-300 also produce nice images but not, IMO, as nice as the E-1 and they lack the live view. I also find the E-300 awkward to use in manual mode, which is how I do nearly all of my shooting. The E-500 and E-330 are much better in that regard.

I agree with you about noise.

Scott

--
Scott
http://smwhittemore.smugmug.com/

 

[Jay Turberville]

refering to "my hypothesis".

I don't have a hypothesis, so you are either not reading or not
grasping.

Actually you do. Your very first statement was, "I've long got the feeling that CCD was nicer than CMOS, and by extension NMOS, but not by much."

Personally, that raised an eyebrow for me as well. My observation is that the character of a camera's images usually has more to do with other factors. For instance, Nikon D2X images look a bit different than images from a 5D. I raise the same eyebrow when I hear people say that the E-1, SD-10, and film all have a "3D" look.

Factors such as Canon's on-chip noise reduction scheme, analog processing approaches, filter choices, well capacities and other design characteristics would seem likely to me to have a greater chance at having a visual impact than the process.

http://web.canon.jp/Imaging/cmos/technology-e/noise_reduction.html

What I'm saying is that E330 is not producing shots with a "Canon"
look to them.

And since the NMOS sensor doesn't use (as far as I know) a scheme like Canon's to equalize amplifier noise, I don't see any reason to expect that it would.

Probably the one thing we can all agree on is that any differences that do exist between sensors tend to be small and typically require close examination to reveal. As such, these differences can often be accounted for by different image processing.

When I saw that the E-330 NMOS sensor used a scheme to give a sensor well size about the same size as found on a similar CCD, my assumption was that the sensors would be similar. My biggest concern wasn't about CCD, but was about Panasonic - who has a reputation of makind CCDs that are noisier than Sony's.

Either way, I do agree that the E-330 makes perfectly nice images. Certainly as good as the E-300 and E-500. My tests show that the DR from all three cameras is essentially the same (using Imatest and the 0.25 med-high as a threshold). Of course, such tests don't account for subjective differences in noise pattern or character.

--
Jay Turberville
http://www.jayandwanda.com

 

[Jay Turberville]

As to the more philosophical point, I AM arrogant. 25 years of
being paid to advise people on technology has taught me that 50% of
people refuse to think rationally, indeed hold rationality in
contempt, of the remaining 50%, 50% are too poorly educated to
grasp the issues, and of remainder, 50% are simply too thick to
have anything explained to them. I have a rather jaundiced view of
the human race...

My experience is a bit different. I observe that thinking logically/rationally takes effort. The natural minute-to-minute human thought processes are heavily oriented toward repetition and pattern matching. And its a good thing to. Logical/rational thinking is much too slow.

The result is that most people are reluctant to take the effort to really think things through unless there is strong motivation. Its not that they can't figure stuff out. Its that there doesn't seem to be enough reward for doing so. That's why advertising works.

Now you say that people are thick. I disagree. I say they just have different priorities. If anything, they are lazy. I've seen too many people that I thought were just average smart or less become quite brilliant when motivated or surprise me with their in-depth knowledge or ability in some surprising area.

In many ways, the average person is often much more time-efficient than so-called "analytics" like me and you. The simple fact is that the 5% of the world that tends toward being analytical combine with the machinations of marketplace to make it hard to find a bad digital camera. Furthermore, the fact that most of these people don't want to be bothered learning the ins and outs of photograph (not that they couldn't if they wanted to) has been a major factor in bringing all the camera automation that we currently have.

This group of people that you call "thick" get most of what they want with a minimal amount of effort. They understand through experience how things work and act accordingly. It isn't necessarly to analyze the heck out of it the "analytics" are willing the shoulder that burden.

Sure, there are potential pitfalls for these folks. But they can usually return the item or sell it on ebay if it doesn't suit them. But for the vast majority of people things work out pretty well most of the time with little effort. That doesn't seem "thick" to me.

So everybody is entitled to their opinion, but I don't have to
"respect" it - in fact, if it irrational, ill thought and dim, I'll
sneer openly at it. And if they feel the same way about me, fine,
I don't give a stuff.

So you sneer unprovoked at the average person and calling them "thick."

Brilliant!

--
Jay Turberville
http://www.jayandwanda.com