Deep red rendering on CMOS and CCD

tr13 · 11-20-2009, 02:03 AM

Just compared CCD and CMOS rendering of red velvet roses.
You don't need to be expert to see deeper CCD image without such blown reds (and historgams are pretty the same). In fact both are quite far from reality (and exposure or any other play don't fix this either).
Purple and deep red are always difficult for digital sensors.

Looks like this:

Attached Images

wlachan · 11-20-2009, 02:21 AM

No conclusion can be drawn I am afraid.

**Ash** · 11-20-2009, 02:45 AM

I see that in my own results, probably not so much here.
But there's a richness to CCD colours that I prefer over the CMOS rendering.
This is what keeps me coming back to the K10D.

tr13 · 11-20-2009, 03:07 AM

Originally posted by wlachan

No conclusion can be drawn I am afraid.

I agree on my small IPS monitor difference is not so much noticeable and roses are mostly dark red (but it is best shot from GX-20 with EV compensation and after slightly playing with color versus random auto shot of cheap TZ5).
On large Wide Gamut monitor difference is much more easeally noticeable (and colors are much closer to reality). Purple is worse on CMOS and rose just becomes red versus richer CCD color.

amoringello · 11-20-2009, 03:57 AM

There is not much to note from the comparison of these two images.

First thing you need to do is keep a consistent environment. Same exposure, same lens, same settings. You clearly have different f-stop, and from what I can see the flower was moved in some way... so perhaps lighting is different. (especially note the thin green stem sticking up along the left side) It is possible you even used different lenses which will have different color characteristics.

I do not doubt you, as I have no experience with the CMOS sensors yet... but I would be interested in seeing a truer comparison.

I am waiting for an excuse to get a K7, but too many issues keep coming up that prohibit me from spending any money on that one.

tr13 · 11-20-2009, 04:15 AM

Please, read carefully - I used tz5 compact, so lens is different, of course.
And F1.7 APS-C DOF is hard to compare to F3.3 1/2.5" DOF.

eigelb · 11-20-2009, 11:09 AM

what's the deal ?

mini photos
different DOF
different EV

**Lowell Goudge** · 11-20-2009, 11:16 AM

Originally posted by tr13

Please, read carefully - I used tz5 compact, so lens is different, of course.
And F1.7 APS-C DOF is hard to compare to F3.3 1/2.5" DOF.

unfortunately you did not mention this in your post and I do not see any of this in the visible exif, so how are we to draw a conclusion between a DSLR and a compact camera?

tr13 · 11-20-2009, 11:23 AM

Originally posted by Lowell Goudge

unfortunately you did not mention this in your post and I do not see any of this in the visible exif, so how are we to draw a conclusion between a DSLR and a compact camera?

I didn't menton this because it is not important here.
And I am pretty sick from DSLRs snobs, all I want to say is that compact colors can be beter, but both are very far from visible image.

Eruditass · 11-20-2009, 12:57 PM

Attacking people isn't the best way to get your point across, even when someone asks a question in a rude way.

I don't know about the compact, but you can set the DSLR to use sRGB vs AdobeRGB or ProPhotoRGB if viewing on a wide gamut calibrated screen and compare that, along with RAW.

To me, the DSLR definitely looks worse and appears to be clipping some red, which usually happens in the default "bright" setting vs natural.

Also, the JPEG compression artifacts appear worse on the DSLR at the top of the rose.

I agree with your assessment of these particular pictures but would like to see more tests done before drawing any conclusion.

Kirivon · 11-20-2009, 01:13 PM

I honestly don't understand what you're trying to prove here.

The K10D with its CCD sensor was notorious for blowing out reds, which is pretty much the only reason I shoot in natural mode rather than bright. In fact, when the K20D came out, the CMOS sensor was considered an improvement in this regard.

Also, the dSLR image has noticeably more saturation, not just in the reds but the greens as well, and the white balance appears to be just a slight bit warmer.

amoringello · 11-20-2009, 02:34 PM

Originally posted by tr13

Please, read carefully - I used tz5 compact, so lens is different, of course.
And F1.7 APS-C DOF is hard to compare to F3.3 1/2.5" DOF.

Oh crap. Another troll.

When you're actually ready to learn, feel free to come back.
Doing a little investigation and getting more information about your subject might be helpful before you try to be so argumentative. Instead you simply sound like a fool. :ugh:

Eruditass · 11-20-2009, 02:36 PM

I wouldn't call him another troll, he seems to be spearheading the Pentax firmware hack operation.

amoringello · 11-20-2009, 03:09 PM

Originally posted by Eruditass

I wouldn't call him another troll, he seems to be spearheading the Pentax firmware hack operation.

A troll by any other name is still a troll. If it smells like a troll, and it acts like a troll, its a troll.

(his comments are argumentative for the sake of argument)

Anyway, I'll have to take your word for his credentials... but for someone who needs to get that in depth with software, he has quite the lack of understanding of how to look at a subject and break it down logically to properly determine how to go about an investigation. This sort of critical thinking should be trivial for someone who is comfortable with reverse engineering firmware or handling such a project.

As I've said, his comments are argumentative for the sake of argument. I would not put a lot of trust in the work coming from such a person. I've been doing software development and quality assurance for the better part of 30 years and I've seen companies fall apart when its leaders have that sort of attitude. I would wish to neither work with nor for such a person.

I wish them all good luck.

ytterbium · 11-20-2009, 03:55 PM

Maybe first start with an analysis how ACTUALLY each of the sensors work - where, why and if at all you should expect such differences to happen.
how many of you have actually seen TRUE raw sensor output directly from ADC?

11-20-2009, 02:03 AM	#1
tr13 Veteran Member Join Date: Jul 2009 Location: Russia Posts: 343	Deep red rendering on CMOS and CCD Just compared CCD and CMOS rendering of red velvet roses. You don't need to be expert to see deeper CCD image without such blown reds (and historgams are pretty the same). In fact both are quite far from reality (and exposure or any other play don't fix this either). Purple and deep red are always difficult for digital sensors. Looks like this: Attached Images

11-20-2009, 12:57 PM	#10
Eruditass Veteran Member Join Date: Jul 2009 Photos: Gallery Posts: 2,207	Attacking people isn't the best way to get your point across, even when someone asks a question in a rude way. I don't know about the compact, but you can set the DSLR to use sRGB vs AdobeRGB or ProPhotoRGB if viewing on a wide gamut calibrated screen and compare that, along with RAW. To me, the DSLR definitely looks worse and appears to be clipping some red, which usually happens in the default "bright" setting vs natural. Also, the JPEG compression artifacts appear worse on the DSLR at the top of the rose. I agree with your assessment of these particular pictures but would like to see more tests done before drawing any conclusion. Last edited by Eruditass; 11-20-2009 at 01:15 PM.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
CMOS vs. CCD and sensor blooming	DonDouglas	Pentax DSLR Discussion	2	07-30-2010 07:29 AM
K10D CCD vs. CMOS K 20D	kleinstein	Pentax DSLR Discussion	9	05-18-2010 12:34 AM
Macro Rendering	Barnster	Post Your Photos!	1	12-31-2009 07:41 AM
Is the new Cmos sensor as sensitive as CCD?	ChrisJ	Pentax DSLR Discussion	2	06-29-2009 11:45 PM
CCD Vs CMOS	MikePerham	Pentax DSLR Discussion	20	08-13-2008 05:39 PM

11-20-2009, 02:21 AM	#2
wlachan Veteran Member Join Date: Sep 2006 Location: Canada Photos: Albums Posts: 2,625	No conclusion can be drawn I am afraid.

11-20-2009, 02:45 AM	#3
Ash Loyal Site Supporter Join Date: Mar 2007 Location: Toowoomba, Queensland Photos: Gallery \| Albums Posts: 23,920	I see that in my own results, probably not so much here. But there's a richness to CCD colours that I prefer over the CMOS rendering. This is what keeps me coming back to the K10D.

11-20-2009, 03:07 AM	#4
tr13 Veteran Member Join Date: Jul 2009 Location: Russia Posts: 343 Original Poster	Originally posted by wlachan No conclusion can be drawn I am afraid. I agree on my small IPS monitor difference is not so much noticeable and roses are mostly dark red (but it is best shot from GX-20 with EV compensation and after slightly playing with color versus random auto shot of cheap TZ5). On large Wide Gamut monitor difference is much more easeally noticeable (and colors are much closer to reality). Purple is worse on CMOS and rose just becomes red versus richer CCD color.

11-20-2009, 03:57 AM	#5
amoringello Veteran Member Join Date: May 2008 Location: Virginia, USA Photos: Gallery Posts: 1,562	There is not much to note from the comparison of these two images. First thing you need to do is keep a consistent environment. Same exposure, same lens, same settings. You clearly have different f-stop, and from what I can see the flower was moved in some way... so perhaps lighting is different. (especially note the thin green stem sticking up along the left side) It is possible you even used different lenses which will have different color characteristics. I do not doubt you, as I have no experience with the CMOS sensors yet... but I would be interested in seeing a truer comparison. I am waiting for an excuse to get a K7, but too many issues keep coming up that prohibit me from spending any money on that one.

11-20-2009, 04:15 AM	#6
tr13 Veteran Member Join Date: Jul 2009 Location: Russia Posts: 343 Original Poster	Please, read carefully - I used tz5 compact, so lens is different, of course. And F1.7 APS-C DOF is hard to compare to F3.3 1/2.5" DOF.

11-20-2009, 11:09 AM	#7
eigelb Inactive Account Join Date: Mar 2009 Location: Italy Posts: 174	what's the deal ? mini photos different DOF different EV

11-20-2009, 11:16 AM	#8
Lowell Goudge Site Supporter Join Date: Jan 2007 Location: Toronto Photos: Gallery \| Albums Posts: 17,886	Originally posted by tr13 Please, read carefully - I used tz5 compact, so lens is different, of course. And F1.7 APS-C DOF is hard to compare to F3.3 1/2.5" DOF. unfortunately you did not mention this in your post and I do not see any of this in the visible exif, so how are we to draw a conclusion between a DSLR and a compact camera?

11-20-2009, 11:23 AM	#9
tr13 Veteran Member Join Date: Jul 2009 Location: Russia Posts: 343 Original Poster	Originally posted by Lowell Goudge unfortunately you did not mention this in your post and I do not see any of this in the visible exif, so how are we to draw a conclusion between a DSLR and a compact camera? I didn't menton this because it is not important here. And I am pretty sick from DSLRs snobs, all I want to say is that compact colors can be beter, but both are very far from visible image.

11-20-2009, 01:13 PM	#11
Kirivon Senior Member Join Date: Sep 2007 Location: Colorado Posts: 173	I honestly don't understand what you're trying to prove here. The K10D with its CCD sensor was notorious for blowing out reds, which is pretty much the only reason I shoot in natural mode rather than bright. In fact, when the K20D came out, the CMOS sensor was considered an improvement in this regard. Also, the dSLR image has noticeably more saturation, not just in the reds but the greens as well, and the white balance appears to be just a slight bit warmer.

11-20-2009, 02:34 PM	#12
amoringello Veteran Member Join Date: May 2008 Location: Virginia, USA Photos: Gallery Posts: 1,562	Originally posted by tr13 Please, read carefully - I used tz5 compact, so lens is different, of course. And F1.7 APS-C DOF is hard to compare to F3.3 1/2.5" DOF. Oh crap. Another troll. When you're actually ready to learn, feel free to come back. Doing a little investigation and getting more information about your subject might be helpful before you try to be so argumentative. Instead you simply sound like a fool. :ugh:

11-20-2009, 02:36 PM	#13
Eruditass Veteran Member Join Date: Jul 2009 Photos: Gallery Posts: 2,207	I wouldn't call him another troll, he seems to be spearheading the Pentax firmware hack operation.

11-20-2009, 03:09 PM	#14
amoringello Veteran Member Join Date: May 2008 Location: Virginia, USA Photos: Gallery Posts: 1,562	Originally posted by Eruditass I wouldn't call him another troll, he seems to be spearheading the Pentax firmware hack operation. A troll by any other name is still a troll. If it smells like a troll, and it acts like a troll, its a troll. (his comments are argumentative for the sake of argument) Anyway, I'll have to take your word for his credentials... but for someone who needs to get that in depth with software, he has quite the lack of understanding of how to look at a subject and break it down logically to properly determine how to go about an investigation. This sort of critical thinking should be trivial for someone who is comfortable with reverse engineering firmware or handling such a project. As I've said, his comments are argumentative for the sake of argument. I would not put a lot of trust in the work coming from such a person. I've been doing software development and quality assurance for the better part of 30 years and I've seen companies fall apart when its leaders have that sort of attitude. I would wish to neither work with nor for such a person. I wish them all good luck.

11-20-2009, 03:55 PM	#15
ytterbium Veteran Member Join Date: Jan 2008 Posts: 1,076	Maybe first start with an analysis how ACTUALLY each of the sensors work - where, why and if at all you should expect such differences to happen. how many of you have actually seen TRUE raw sensor output directly from ADC?