The Future of CCD Sensors

[Atina]

With the introduction of CMOS sensors into the medium-format world, what does the future look like for CCD sensors?

In your view, is there a CCD look or is it something that is vastly overvalued?

Is CMOS in your future or are you one of the photographers who is not so impressed with the latest CMOS medium-format sensor offerings?

What is your view in general about the impressions of other photographers about the CMOS sensors? Is the problem only the price? Do you predict a good future or a serious flop of the new equipment?

[fotagf8]

I understand the advantages of CMOS sensors, particularly if you are shooting sports or street photography or concerts, but for lots of landscape and architectural photography, I am not sure that those advantages are all that important.  To the extent medium format cameras are larger than DSLRs, the camera is likely an impediment in street and sports photography.  There is also the question of whether using the camera hand-held (which becomes more likely with a CMOS sensor) will be satisfactory to photographers who want sharp photos without visible camera shake.

[BJL]

I understand the advantages of CMOS sensors, particularly if you are shooting sports or street photography or concerts, but for lots of landscape and architectural photography, I am not sure that those advantages are all that important.

Putting aside some very bogus comparisons that mix together many other differences along with the sensor type, this seems to be the only remaining argument against switching from CCD to CMOS: that its advantages are less important in some case than others.  But with some clear upsides to CMOS and no real downsides, the change seems inevitable, and the recent adoption of CMOS by Leica, Phase One and Hasselblad along with the total abandonment of CCDs in all other systems in 35mm format and smaller should make it clear which way the wind is blowing.

What is more, some advantages of modern CMOS sensors are useful also for some landscape and architectural photography, such as increased dynamic range (through far lower dark noise) and magnified live view for tasks like careful manual focusing at off-center points.

P. S. As other posts have mentioned, I overlooked one current advantage of CCDs: being available in sizes larger than 36x24mm 44x33mm.  I will not speculate if and when a CMOS sensor maker like Sony will venture into that territory; it might be that the larger formats will remain "too niche" for CMOS (which seems to require larger production volumes to cover higher fixed costs like R&D), which could mean that the largest format "low ISO speed tripod cameras" stay with CCDs.

[ErikKaffehr]

Hi,

It doesn't seem that CCDs have any advantage over CMOS, possibly except astronomical applications where they are cooled with liquid nitrogen.

Well, except they are made in larger sizes… Also, it may be that they may work better with large beam angles, but that may also change.

Best regards
Erik

With the introduction of CMOS sensors into the medium-format world, what does the future look like for CCD sensors?

In your view, is there a CCD look or is it something that is vastly overvalued?

Is CMOS in your future or are you one of the photographers who is not so impressed with the latest CMOS medium-format sensor offerings?

What is your view in general about the impressions of other photographers about the CMOS sensors? Is the problem only the price? Do you predict a good future or a serious flop of the new equipment?

[sbernthal]

I don't believe there is a real difference in colors between CCD and CMOS.
CCD has an advantage right now in having very large sensors available.
If they could make 100MP 53x40mm CMOS sensors at reasonable cost, I don't think they would be introducing new CCD backs after that.

[EricWHiss]

The differences may well chalk up to design directives not what is in theory possible with each platform.  What I mean is it seems most CMOS based sensors are designed with the primary goals of achieving higher ISO capabilities plus video or at least live view.    CCD sensors seem to have goals of achieving the best possible IQ at base ISO. All IMHO of course.   

When I look at the full res samples from the pentax 645z and the IQ 250, I really don't think the same crispness at the pixel level is achieved as with the CCD sensors but I can't say why.    I don't agree with others who say that these CMOS sensors are better files in every way.  I see pluses and minuses.  The big plus is of course the video possibility and the few stops more usable ISO.   

[Jim Kasson]

It doesn't seem that CCDs have any advantage over CMOS, possibly except astronomical applications where they are cooled with liquid nitrogen.

Well, except they are made in larger sizes… Also, it may be that they may work better with large beam angles, but that may also change.

The large size is a reason that CCDs are used in large astronomical telescopes, since there's a limit to how small the sensels can be even without regard to their light-gathering capacity. Atmospheric turbulence causes distant point sources to spread to 0.4 arc-second disks at even the best locations. For a 2-meter f/6 telescope, that means that pixels smaller than 12 microns on a side won't help the resolution. Thus the way to get decent resolution is to make the chips, or chip assemblies, big. A 4096x4096 chip with a 12-micron pitch would have an active surface that would measure almost 2 inches on a side.

Here's a 4Kx4K chip used in astronomy that uses a 15 micron sensel pitch.

http://www.fairchildimaging.com/catalog/focal-plane-arrays/ccd/ccd-6161

OK, I admit it's off-topic, but it's kinda fun.

Jim

[Telecaster]

I'm inclined to buy into the idea that Color Filter Array differences have more impact on perceived image quality & tonality than inherent differences between CCD & CMOS sensors. I will say that my prefered digital-era tonality comes mostly from cameras with CCDs. Nikon D70, Pentax K100D & Epson R-D1 (all using variants of the same Sony 6mp sensor), Nikon D80 & Pentax K10D (same 10mp sensor) and more recently a Pentax 645D and secondhand Leica M8(.2). These cameras all produce RAW files with distinct hues as opposed to the more "global" (as BCooter aptly puts it) color of many/most other cameras. My guess is the CFAs used in these cameras are relatively narrowband in their light transmission. It is possible to get this look with CMOS too...the Olympus E-M5 is an example.

Last week I did a comparison between my Sony A7r and M8, a Zeiss 21mm on the Sony and a Voigtländer 15mm on the Leica (for a near-identical field-of-view). Various outdoor scenes in my neighborhood. In terms of sheer detail the Sony comes out on top, of course (though the M8's per-pixel resolving capability is impressively high), but tonally the Leica has a clarity & separation the Sony lacks. I can make the Sony look kinda like the M8 in post, but its files lack that extra bit of chromatic zing. I suspect the A7r may soon end up with an acquaintance while the M8 may soon be joined by a backup.

-Dave-

[sbernthal]

Where can you see ISO 100 short exposure files from IQ250?

I don't think Pentax is the right comparison, and high iso and long exposure files should not be the base to compare the best that that back can do.

[Doug Peterson]

Where can you see ISO 100 short exposure files from IQ250?

I don't think Pentax is the right comparison, and high iso and long exposure files should not be the base to compare the best that that back can do.

As a dealer we maintain a large library of sample raw files for our customers to use for evaluation, and for our own research/knowledge. I imagine other dealers do as well, but cannot speak on their behalf. I have IQ250 raw files from the weddings I've shot it, our tech camera testing, several customer tests (with their permission of course), and some technical comparisons we've done.

[sbernthal]

In the link that you provided the downloads are dead.

Can you please provide other links to pages where you can download raw files of ISO 100 images with full studio lighting?

The only short exposure ISO 100 image I could find did not look very good.

[ErikKaffehr]

Hi,

Technically both CCD and CMOS are monochrome devices collecting electrons. The major parameters are what is called full well capacity (FWC), the number of electrons a sensel can hold and the readout noise. So CCD or CMOS has no effect on colour at all. Colour rendition is determined by the CFA array and some math.

FWC determines noise at medium exposure levels, and that noise is almost entirely dependent on statistical distribution of light.

CMOS can do some tricks:

1) It can measure remaining charge after reset and before exposure and subtract it from recorded exposure. It is called correlated double sampling. All CMOS sensors do that.

2) Many sensors use on chip converters, 6000-8000 of them. These can work in parallel and therefore they can have long integration times. Sony sensors have this and so does the CMOSIS chip on Leica. On chip converters seem to have very low noise. Many cameras use Sony sensors and other firms also have on chip converters. Nikon D3, D3s, D4 and all Canon DSLRs use off chip conversion, giving up significant DR at low ISO and gaining some sensivity.

CFA design varies a lot, this article offers some insight: http://dougkerr.net/Pumpkin/articles/Metameric_Error.pdf

Different makers can build different CFA-s.

The major advantage of CMOS over CCDs is the reduced readout noise, especially with on chip converters. On chip converters simply output numbers, while CCD needs complex readout electronics and of chip ADCs.

Personally, I have experience of both CCD and CMOS devices:

CCD: Minolta Dimage 7D, Sony Alpha 100 and Phase One P45+
CMOS: Sony Alpha 700, 900, 55, 77, and 99

Regarding colour rendition, it may be highly personal. What I can say that there is not a lot of difference between my most used cameras right now, the Sony Alpha 99 and the P45+ if both are processed using the same raw processor using DNG profiles generated the same way and identical white balance.

So, what I can say that there is little technical explanation of any advantage of CCDs over CMOS, and this is consistent with my own experience.

But, I am an engineer and no great artist…

Best regards
Erik

I don't believe there is a real difference in colors between CCD and CMOS.
CCD has an advantage right now in having very large sensors available.
If they could make 100MP 53x40mm CMOS sensors at reasonable cost, I don't think they would be introducing new CCD backs after that.

[Doug Peterson]

In the link that you provided the downloads are dead.

Can you please provide other links to pages where you can download raw files of ISO 100 images with full studio lighting?

The only short exposure ISO 100 image I could find did not look very good.

We are in the midst of changing from Dropbox to google drive. Until this is complete the links will remain publicly unavailable. After completing the switch to Google Drive we will have all the previous samples back up as well as several new raw samples. In the meantime our customers can email me and I'll gladly provide a manual share. In the meantime, please contact your local dealer to work with them from their inventory of sample and test files.

I honestly can't imagine what shot from the iq250 you could possibly have seen in raw format and not been impressed.

[michael]

Thanks for your input Eric. I was told by a senior MF back designer a few years ago that CCDs and CMOS don't differ re colour. As you say, it's the CFA and other factors.

Time to put that myth to bed.

Michael

[sbernthal]

I'm not really in the market for a new back right now.
I saw one example of an outside shot jpg that looked pretty Canonny to me. I'm mot sure I can find it again.
But that doesn't interest me - I will wait until somone releases studio shots of ISO 100 still objects, hopefully alongside IQ180/280 ISO35, so we can compare best/best.

[Doug Peterson]

I'm not really in the market for a new back right now.
I saw one example of an outside shot jpg that looked pretty Canonny to me. I'm mot sure I can find it again.
But that doesn't interest me - I will wait until somone releases studio shots of ISO 100 still objects, hopefully alongside IQ180/280 ISO35, so we can compare best/best.

We have done several such comparisons including a rather nice food shoot.

[sbernthal]

Is it available on your website as jpg?

[eronald]

Thanks for your input Eric. I was told by a senior MF back designer a few years ago that CCDs and CMOS don't differ re colour. As you say, it's the CFA and other factors.

Time to put that myth to bed.

Michael

Michael,
 There are two different questions:
 1. Is *ideal* CMOS different from ideal CCD?
 2. Is *real* CMOS different from real CCD.

You say no to 1. I would agree that they are reasonably close - you need to remember that different CCD technologies are actually quite different eg. in the spectral sensitivity of the base process, even before CFA color filtering.

Regarding 2, My impression is that most *existing* CMOS implementations have been more strongly digitally filtered than their CCD counterparts, and the result is bad skin tone (the plastic wrap look). This does not appear to be a basic color issue, objects usually photograph well with the same cameras. Maybe the new Sony CMOS is closer to clean. Obviously, CMOS can do a lot more tricks, but a CMOS chip that has warmed from normal use of liveview is going to be much noisier than a cold CCD that *cannot* do liveview... in the real world.

People used to shoot Ektachrome or Kodachrome depending on the look they wanted, I see no reason to assume one size will suddenly fit all just because the physics of capturing a momentary slice of light has been shoehorned into direct digital output.  

As to my qualifications - what can I say? I used to be a scientist, but when it comes to images I say lab measurement should be used to confirm and explain subjective impression rather than replace it. Color is not a physical entity in its own right, it is an evoked sensation.

Edmund

[ErikKaffehr]

Hi,

I guess that some of the "plastic wrap look" may come from the low noise of CMOS.

The Alpha 99 I have has electronic viewfinder and thus uses the sensor full time, but it has significantly less noise than my P45+ in real world outdoor shoots (with matching raw histograms).

On a DSLR the use of live view is optional. Some MF CCD sensors do have live view.

I can't say I can see a lot of difference between P45+ and Sony Alpha 99 in A2 size prints. I can't actually tell them apart on real world subjects. At larger sizes I'm pretty sure that the 39 MP advantage of the P45+ over 24 MP from the Alpha 99 shows.

Best regards
Erik

Michael,
 There are two different questions:
 1. Is *ideal* CMOS different from ideal CCD?
 2. Is *real* CMOS different from real CCD.

You say no to 1.

Regarding 2, My impression is that most *existing* CMOS implementations have been more strongly digitally filtered than their CCD counterparts, and the result is bad skin tone (the plastic wrap look). Maybe the new Sony CMOS is closer to clean. Obviously, CMOS can do a lot more tricks, but a CMOS chip that has warmed from liveview is going to be much noisier than a cold CCD ... in the real world.

Edmund

[sbernthal]

The ideal look is not important.
What is important is the real look under best conditions.
You can also make comparisons under harsh conditions, but then they apply only to those specific conditions.

The models you quoted are not best of breed, so I would not use them to come to a conclusion.
P45 is using the old Kodak sensor, and we know very well that the DALSA sensors produce better images.
We also know that the IQ/Credo produces better images than P/Aptus, so that is what needs to be compared.

Comparing 5-6 years old technologies maybe gives us a picture of what was right then, but we want to know the answer that is applicable today, considering what the relevant companies are actually making.

I believe the plastic wrap look is a result of exaggerated noise reduction, not of naturally low noise in CMOS.
When you are comparing a Canon image to a Leaf image, you are not only comparing CCD to CMOS, but also the two companies very different approaches to noise reduction and general rendering.

If Doug would feel generous, he could release to us a couple of images by IQ CMOS and IQ CCD using controlled lighting and the lowest ISO each back does, and then we will know what is best right now.