CMOS Vs CCD

[Kalus]

Yes its usually because of the heat build up from running it so long (and the temperature of the environment) but a dark frame isn't exactly to get rid of the noise.

It's called noise reduction but all its really good for and primarily used for is to get rid of "hot pixel" noise that comes up in longer exposures while most ppl when they hear noise they think of blue sky/shadow noise etc not hot pixels. If you were to take a long exposure blue sky picture which had noticable noise in it (ala any prosumer camera) then running a dark frame on it wouldn't do too much but just get rid of the noise caused by the dark current in the sensor but most of the visual noise isn't from dark current so you'll still have a noisy blue sky.

Great overview. I would also like to add that many CCD cameras use
a second "dark frame" technique to reduce noise at slow shutter
speeds. basically, it will take a sceond frame, of pure black, to
measure where the noise is and use that to try and eliminate the
noise in the actual photo. Usually this means that CCD cameras
will take much longer to handle long exposures. This is partially
due to the fact that CCDs tend to get noisier the longer they are
left "on." I THINK this is due to heat sensitivity.

--
Ryusen: Stand up Philosopher
http://www.pbase.com/ryusenkai/
ryusenkai.org for some pictures. Early shots taken with Canon G-1,
later stuff with 300D

 

[Mark Marshall]

Even if not, I'm sure it will be selectable. Nikon are twiddly happy from my experience. I'm not sure how it's done on the D100, but if it's menu driven like the Coolpix line then no thanks anyway. ~ m²

2.for those comparing the D70 and 300D, i'd like to remind that, it
doesn't read anywhere about the metering 'selectability' on the
D70, right?

--

'Brothers and sisters, we've learned that there's some bad bokeh going around out there. So like, just be careful man, alright?' (If Wavy Gravy emceed PMA) http://rhodeymark.instantlogic.com

 

[The Krakken]

I was so disgusted with my pics at f2.8 that I couldn't bear to look at them, and as a result, erased them as fast as I could. I sent a sample CD with my lens into Sigma. When it gets back (hopefully in 2 weeks), I'll post them.

Maybe I'll do it anyway, whenever my lens gets back from Sigma.
Wouldn't you know it, that same lens was soft at f2.8???

I sure hope they fix it. The extra stop of light was the only
reason I chose it over the Canon F4L.

--
Ryusen: Stand up Philosopher
http://www.pbase.com/ryusenkai/
ryusenkai.org for some pictures. Early shots taken with Canon G-1,
later stuff with 300D

--

'One more time we hear.......the images look great, just so long as you don't look at them'

Justin Blethrow on the rabid defense of a certain 'other' camera.

http://www.pbase.com/daemeon

 

[Negative287]

True but the noise that is there can be exaggerated in the luminance channel.

The thing I notice most when I look at shots from each sensor is the CCD tends to have more luminance noise (Salt and Pepper type) while the CMOS tends to have more chrominance noise (color channel) or they appear that way even though they may be more even then anyone realizes.

It may or it may not, theres 2 reaons CCD's are more sensitive to
light the CMOS.

good or bad?...does this mean it's more susceptible to noise?

--

'The only real currency in this bankrupt world is what we share with each other when we're being uncool.' -- Cameron Crowe

 

[Negative287]

Here's the problem... Image was as clean full size as you were seeing it reduced... It is also resized simply because that's the way I view it full screen...

As for the details though the issue is in the way the cameras are processing the files. I find that the Nikon sharpening is a typical sharpening of every pixel (I forget the name for that type of sharpening) and the Canon sharpening tends to be more of edge sharpening.

When I shoot NEF using the Nikon I can just set PS Plugin Up to handle the sharpening and NR (The Nikon doesn't implement any NR on NEF unless it is long exposure) the way that I want and then I get a similar look to what the 10D offers (well usually I use less NR and more sharpening but that is just my preference).

The truth is I really don't think this deal is as big as people made it and I still contend that a lot of Phil's tests are Point and Shoot tests but I guess people need a system by which to measure so his is as good as any.

Pete Perry wrote:
[snip]
I agree with you Pete, but the image you linked is a resize - not
really a fair comparison to the other one.

Personally having seen both full size samples from both, the
difference is noticable enough for me to stick with Canon's sensor
for the moment.
--
http://www.pbase.com/timothyo

--

'The only real currency in this bankrupt world is what we share with each other when we're being uncool.' -- Cameron Crowe

 

[remi176129]

Actually, I did. I erased them, as I didn't think they were
useful. They were taken with my F717, and were outclassed by the
one's take from my 300D.

It's obvious that a big CMOS will make better pictures than a small CCD.
You can't really compare sonys minute CCDs to the 1.5x one of the D70!

 

[Another birder]

Here's the problem... Image was as clean full size as you were
seeing it reduced... It is also resized simply because that's the
way I view it full screen...

Any chance of a link to a full size version? (or any other full size D100 images you have!) I would be more than happy to put the CMOS vs CCD issue 'to bed' in my own mind.

The truth is I really don't think this deal is as big as people
made it and I still contend that a lot of Phil's tests are Point
and Shoot tests but I guess people need a system by which to
measure so his is as good as any.

True - I can only judge the quality differences from sites like dpreview, and they give the edge to the CMOS sensor to my eyes. (not by a huge amount in most cases, but enough for me to list it as an advantage for Canon over Nikon in my mind)
--
http://www.pbase.com/timothyo

 

[CaryT]

Not to toss any more gas on flames, but check out this 'summary' of technical diffs between the two:

http://electronics.howstuffworks.com/digital-camera3.htm

I looked into this 'cause the EE here at work didn't seem to know about a difference (technically-speaking) between the two. Oddly, this article seems somewhat backwards, when compared to the results we're getting with our DRs...

Maybe it's dated info, maybe it's just bad info? Maybe generalizations are bad (isn't that redundant?)?

-Cary

 

[lamha]

This report, from a college student it seems, is an interesting read.

http://www.eng.auburn.edu/~troppel/General_Interest/CMOS_Imaging/Rama_Written_Report.pdf

first of all..i'm not trying to stir up another 'holy war'
defending 'mine's better!'

second, i do believe there are tons of people thinking of getting a
300D on this forum...thus i consider this is 300d related
topic...not OT

when someone brought up the D70 issue...i believe for those who are
pondering between 300d and D70, two major concerns:

1. pics from CMOS and CCD taste different...according to some
threads i read here, what do you think on this? what kind of
different tastes then?sharpness? expousure tendency? saturation? or
others?

2.for those comparing the D70 and 300D, i'd like to remind that, it
doesn't read anywhere about the metering 'selectability' on the
D70, right?

thanks in advance...your comments are appreciated

 

[Dave Disser]

The biggest difference between CMOS and CCD is that Canon's CMOS chips are much cheaper to manufacture, so don't rush to buy up that Nikon stock. Nikon appears to have made a really nice camera, and people who buy it will be getting a lot of bang for the buck, because Nikon is probably taking a bath on them for the sake of preserving some market share. Look for Canon to counter by dropping the price of the Rebel and pre-announcing a new SLR or two before the D70 ships.

Also, I'll be surprised if we don't hear about some firmware bugs or a schedule slip, as the development teams must have been under tremendous pressure to get something out quick since Canon's delivery of the Rebel last fall was a gut-kick to the entire industry.

--Dave

first of all..i'm not trying to stir up another 'holy war'
defending 'mine's better!'

second, i do believe there are tons of people thinking of getting a
300D on this forum...thus i consider this is 300d related
topic...not OT

when someone brought up the D70 issue...i believe for those who are
pondering between 300d and D70, two major concerns:

1. pics from CMOS and CCD taste different...according to some
threads i read here, what do you think on this? what kind of
different tastes then?sharpness? expousure tendency? saturation? or
others?

2.for those comparing the D70 and 300D, i'd like to remind that, it
doesn't read anywhere about the metering 'selectability' on the
D70, right?

thanks in advance...your comments are appreciated

 

[ryusen]

i shot my original tests at ISO 200 and they were soft at f2.8, but when i shot at iso100 or iso 400, the pics were fine.. i can't figure it out...

I was so disgusted with my pics at f2.8 that I couldn't bear to
look at them, and as a result, erased them as fast as I could. I
sent a sample CD with my lens into Sigma. When it gets back
(hopefully in 2 weeks), I'll post them.

--
Ryusen: Stand up Philosopher
http://www.pbase.com/ryusenkai/

ryusenkai.org for some pictures. Early shots taken with Canon G-1, later stuff with 300D

 

[Negative287]

This is an ISO 400 image and I haven't applied USM to it as of yet... It was Nikon Raw Format and processed using Adobe Photoshop CS Raw Plug-in. But outside of a color profile change nothing was really done to the image in PS itself...

Here's the problem... Image was as clean full size as you were
seeing it reduced... It is also resized simply because that's the
way I view it full screen...

Any chance of a link to a full size version? (or any other full
size D100 images you have!) I would be more than happy to put the
CMOS vs CCD issue 'to bed' in my own mind.

The truth is I really don't think this deal is as big as people
made it and I still contend that a lot of Phil's tests are Point
and Shoot tests but I guess people need a system by which to
measure so his is as good as any.

True - I can only judge the quality differences from sites like
dpreview, and they give the edge to the CMOS sensor to my eyes.
(not by a huge amount in most cases, but enough for me to list it
as an advantage for Canon over Nikon in my mind)
--
http://www.pbase.com/timothyo

--

'The only real currency in this bankrupt world is what we share with each other when we're being uncool.' -- Cameron Crowe

 

[JR79]

The autofocus uses their dedicated sensors (I believe all are CMOS), not the main imaging chip. So this tells nothing.

hmm... that could account for the light sensitivity

D100 EV -1 to 19
D70 EV -1 to 19
DReb EV 0.5 to 18

I wonder if that translates to low light sensitvity for the D70..

--
http://gallery29564.fotopic.net/

 

[Another birder]

This is an ISO 400 image and I haven't applied USM to it as of
yet... It was Nikon Raw Format and processed using Adobe Photoshop
CS Raw Plug-in. But outside of a color profile change nothing was
really done to the image in PS itself...

Thanks Pete. Looks very similar to what I get from the CMOS at 400 - perhaps slightly more detailed actually. (CMOS seems to lump details a bit more at 400 and above)

Cute cat btw.
--
http://www.pbase.com/timothyo

 

[Negative287]

I know it is about the same that I get from the 10D as well and that's why I don't believe that using default settings is ideal or even gives you a real idea of what these two cameras can do.

I'm kind of disappointed because it appears that people who want to just pick the camera up point, click and print are most happy with the S2 and I just can't bring myself to go that direction because that camera has too many other short comings. So, for now I get really good at post processing which is more beneficial anyway.

This is an ISO 400 image and I haven't applied USM to it as of
yet... It was Nikon Raw Format and processed using Adobe Photoshop
CS Raw Plug-in. But outside of a color profile change nothing was
really done to the image in PS itself...

Thanks Pete. Looks very similar to what I get from the CMOS at 400

• perhaps slightly more detailed actually. (CMOS seems to lump

details a bit more at 400 and above)

Cute cat btw.
--
http://www.pbase.com/timothyo

--

'The only real currency in this bankrupt world is what we share with each other when we're being uncool.' -- Cameron Crowe

 

[Medic Laviña]

It depends on who manufactures the sensor. Many DSLRs using CMOS were successful in eliminating the image noise.
Try comparing the 300D and Fuji S2 Pro samples at Steve's site:

300D sample:
http://www.steves-digicams.com/2003_reviews/300d/samples2/CRW_0231.jpg

Fuji S2 Pro sample:
http://www.steves-digicams.com/2002_reviews/s2/samples/DSCF0207.JPG

So far, only the Fuji DSLR has the noise free CCD I know. If you are comparing these sensors (CMOS vs. CCD) just because Nikon announced the D70, I believe Nikon have to do way better than what they did with their D100. Take a look at this sample:

Nikon D100 sample:
http://www.steves-digicams.com/2002_reviews/nikon_d100/samples/DSC_1364.JPG

One good advantage of a CCD sensor (if I'm not mistaken) is the possibility of enabling live LCD preview (or even movie clips capability) on your DSLR. That is, if the manufacturer decides to include those features.

Medic

 

[Kalus]

No thats not a possability, you can do it with CMOS too but most DSLRs still use a mirror to send the optical data to the view finder so a live preview isn't possible unless they flipped the mirror up then activeated it but I don't see why anyone would bother as a main reason you get an SLR is for the TTL optical viewfinder.

A good advantage of CCD's is faster shutter speeds as you can time the CCD on and off after hte shutter is open to get even short shutter speeds ala the 8000 speeds etc while you can't do this with a CMOS (as far as I know)

One good advantage of a CCD sensor (if I'm not mistaken) is the
possibility of enabling live LCD preview (or even movie clips
capability) on your DSLR. That is, if the manufacturer decides to
include those features.

Medic

 

[gerdg]

Once there were SLRs with semi-tranparent fixed mirrors (that wouldnt flip up)

may be that could be a solution for preview and movie with a DSLR !! what do you think, Kallus?
Regards
Gerard

A good advantage of CCD's is faster shutter speeds as you can time
the CCD on and off after hte shutter is open to get even short
shutter speeds ala the 8000 speeds etc while you can't do this with
a CMOS (as far as I know)

One good advantage of a CCD sensor (if I'm not mistaken) is the
possibility of enabling live LCD preview (or even movie clips
capability) on your DSLR. That is, if the manufacturer decides to
include those features.

Medic

 

[Medic Laviña]

No thats not a possability, you can do it with CMOS too but most
DSLRs still use a mirror to send the optical data to the view
finder so a live preview isn't possible unless they flipped the
mirror up then activeated it but I don't see why anyone would
bother as a main reason you get an SLR is for the TTL optical
viewfinder.

A good advantage of CCD's is faster shutter speeds as you can time
the CCD on and off after hte shutter is open to get even short
shutter speeds ala the 8000 speeds etc while you can't do this with
a CMOS (as far as I know)

Well, Olympus already did the live LCD preview on their old E-20.

One main reason to use the LCD (preferrably the swivel type) is when taking shots wherein the use of viewfinder is hard (or impossible). An example is taking an overhead shot in a crowded area, or taking a floorlevel shot that peeking onto viewfinder is kind of awkward.

Medic

 

[Medic Laviña]

No thats not a possability, you can do it with CMOS too but most
DSLRs still use a mirror to send the optical data to the view
finder so a live preview isn't possible unless they flipped the
mirror up then activeated it but I don't see why anyone would
bother as a main reason you get an SLR is for the TTL optical
viewfinder.

A good advantage of CCD's is faster shutter speeds as you can time
the CCD on and off after hte shutter is open to get even short
shutter speeds ala the 8000 speeds etc while you can't do this with
a CMOS (as far as I know)

Well, Olympus already did the live LCD preview on their old E-20.
One main reason to use the LCD (preferrably the swivel type) is
when taking shots wherein the use of viewfinder is hard (or
impossible). An example is taking an overhead shot in a crowded
area, or taking a floorlevel shot that peeking onto viewfinder is
kind of awkward.

BTW, I'm not saying Olympus E-20 is a true DSLR. My point here is live LCD preview can be done even with TTL viewfinder. I believe DSLR manufacturers can do it if they want to.

Medic