HELP! - HC1 Poor Red Color Accuracy

[Alex Thames]

Well, it seems that the A1/C1 red problem is not on all models. Some seem to fare well, others not well at all. The demo model of the A1 at the BH store seemed to reproduce red decently on the LCD screen as well as the Sony monitor it was hooked up too.

[Fredrik-Larsson]

Yeah, it seems as if there are a couple that has bad red. Mine is pretty OK and some clips can be downloaded at:
http://www.apollologic.com/movies/hc1_reds/

EDIT: Whoops, I had already mentioned my reds-clips. Sorry, it's early. :)

[Laurence Kingston]

Originally Posted by John McManimie

While deciding whether to purchase the HDR-HC1 recently (which I did), I read quite a bit regarding CMOS sensors. From my reading, I thought (perhaps mistakenly) that CMOS sensors were actually more sensitive to infrared (IR) than visible light and thus have IR cut filters (as do CCDs as well) to block wavelengths beyond the normal visible spectrum and allow better color performance in a range of lighting conditions (but the IR filter would also unfortunately reduce the camera sensitivity). The filter controls the sensor’s response to the end of the red spectrum and will react differently depending on the intensity of IR (cloudy day versus bright sunny day or room lit by a light bulb).

I have seen posts on the web where others have tested their CMOS based digital cameras to see if near IR sensitivity is affecting color reproduction. One person’s suggestion was to photograph subjects that reflect the long red to near IR region to see if a problem exists (deep blue flowers reproducing purple, deep purple reproducing red violet, and some green pool tables showing up brownish).

I guess that I never expected the camera to be perfect or to provide fixed colors under all conditions (film cameras don’t and our eyes don't either), but to provide consistent color results under the same conditions (and it does). I figured that a manual white balance would be preferable to adjusting saturation and that I would fix color "problems" in post (By the way, I have not seen any problems with red so far).

This post made me think. I have not noticed a problem with reds and since being alerted to this I've really looked for it. I also always shoot in full automatic so you would think that that would make the problem worse. On the other hand, I always have a UV filter in front of the lens, mostly there as a lens protector. Is it possible that extra light beyond what we can see in either the high or low frequency range might be causing this problem that some of you are having? Maybe extra the UV stuff that I'm filtering out is causing this? If the CMOS was capturing UV and interpreting it as violet and mixing it back in with the reds, maybe you would get this effect. I know I don't seem to have this problem, I have an early A1, and I have never really shot without a UV filter.

[Fredrik-Larsson]

Odd... I have also a UV-filter on mine. Some of the first pictures I took was of a red/pink flower and I thought that it wasn't quite right color. I am almost sure that I didn't have an UV-filter back then. Now I always have it on. It might be interesting to try without and compare the photos.

[John McManimie]

Originally Posted by Christian de Godzinsky

...This IS a BIIIIIG disappointment. The reds are very blueish especially under artificial (fluorescent and bright halogen) lighting... It is bareable under a clear sky a bright day...

Originally Posted by Steve Mullen

This is either Sony's estimate of what folks want -- all their cameras are too blue for me -- OR it is inherent to the design.

Perhaps UV is indeed causing additional blue cast to the colors. I saw that very topic discussed here (note the Reader's Comments): http://www.photo.net/equipment/filters/

[Christian de Godzinsky]

Hi,

Now after having used the HC1 for some more serious work I have to say that the red color reproduction is SOO BAD that I would like to let go of this camera. I will NEVER again buy a piece of imaging equipment without checking it out first -personally. From any manufacturer.

I have compared the behaviour of the HC1 now with other prosumer and even consumer DV-cams. They all perform ok compared to the HC1 in the red ... orange region. The difference is HUGE. None of them produce lilac colors when shooting pure red objects - the HC1 does!!!

I am now sending the HC1 to the local Sony service. I asked if they recognize this problem but never got a clear answer. Either they do or they don't, we will never know - but they will have a "look" at it.

This "cam-roundtrip" (all the way to France) is going to take at least two weeks. Probably much more. I will report the results (if any) when I get the camera back.

DOES ANYBODY KNOW if there is a way to check the firmware version of the HC1??? It would be nice to compare if they have updated it, especially if the colors are better after the servicing. That would probably be reassuring to know for anyone out there with the same problem.

And HOW do I earn the priviledge to post attachments in this forum? Would be convenient to send jpegs or small clips for comparison...

REgards,

Christian

[Fredrik-Larsson]

Christian, it probably take a thousand posts or more to be able to do that. But if you don't have a website of your own you can mail me and I will upload them on mine. Being a HC1-owner it's very interesting to learn more and discuss this issue.

[Stu Holmes]

Originally Posted by Christian de Godzinsky

I asked if they recognize this problem but never got a clear answer. Either they do or they don't, we will never know - but they will have a "look" at it.

Hope you get the camera fixed Christian. If it IS fixed, then it would be interesting to know WHAT was changed on the cam - whether it's firmware version or some other re-calibration.

As for the Sony centre being, ah, vague about recognition of it, this is probably their policy not to 'admit' since it may constitute some form of legal liability and open floodgates for other HC1 owners to try to force them into a free fix. That sort of thing - just a guess.

Keep us in the loop as to whether you achieve a food result. Certainly from Steve's (recent poster) experiences, it seems that this red problem IS either fixable, or more accurately, not as apparent on more recent units.

I'm sure that Sony are conscious of it, but playing a little 'dumb' deliberately...

[Steve Mullen]

For those who have this problem -- what happens when you shoot something that's saturated red and:

1) Use WB shift to bias toward green?

2) Use AE shift to bias darker?

3) Also, if you are shooting outdoors -- try fixed OUTDOORS.

4) Also, if you are shooting indoors -- try fixed INDOORS.

5) Try One Touch on a white towel.

Can someone try these tests and report back. In my DV experience, this is an overexposure issue. But this could be different.

[Christian de Godzinsky]

Hi,

And thanks Steve for your comments !!!

Believe me -I have tried both manual and automatic WB, tweaking the color setting and WB shift. So - under some circumstances you CAN get the red more red BUT at the same time you loose the quality of the other colors :(

Adding +2 ..+3 to the color setting seems to help a LITTLE, especially when the extra gain is needed (under lower light conditions) - BUT results in too saturated colors under good lighting. It is not acceptable to tweak such a control every time the lighting condition changes. And as I said - it just helps a little. It does NOT solve the problem.

I have been looking at stills exported from the HC1 using Photoshop and it's tools. The red color that turn lilac is not even close to saturation, there is just too much of blue in the red object!!!

The red and blue colors are spectrally far from each other and should be filtered quite efficiently by the BAYER RGB filter on top of the CMOS sensor. This leaves only a few options, there is something wrong with the color matrix and encoding - probably not fixable swapping the firmware in the control processor. There is something wrong with the CMOS chip itself - or the color filter on top of it.

I am very sceptic about if this can be fixed but I am, naturally, very hopeful :) It seems that my cam really behaves badly. I will set up some pics on my homepage during the weekend so you all can see for yourselves...

Christian

[Steve Mullen]

Originally Posted by Christian de Godzinsky

I have been looking at stills exported from the HC1 using Photoshop and it's tools. The red color that turn lilac is not even close to saturation, there is just too much of blue in the red object!!!

This is the way cameras looked a decade ago. It seems to be either an error in the camera OR a problem when NTSC/PAL is generated.

1) Are you viewing on an HD monitor?

2) Have you tried reducing exposure?

[Dave Halliday]

I can back christian up here. I can also back up most of the other people posting here.

Let me explain. I think the red balance is an issue--there seems to be a very fine margin of error on white balance/color saturation/exposure.

When you see guys like Christian and I getting frustrated with it, its because we're used to cameras like the PD170, etc. working *acceptably* in all circumstances on full Auto. With the A1U you have to watch out more, make more minor adjustments (which is hard b/c the buttons are small) and just generally "work harder" to get a decent shot.

I've tested two other HC1 units at local best buys and the results have been the same: In some cases, the Reds shift to a garish pink. kicking the ae up a stop only makes the shot underexposed. frustrating, but not impossible to work around. Either way, it's better than the HD10 ;-)

[Steve Mullen]

Originally Posted by Dave Halliday

Either way, it's better than the HD10 ;-)

I hope you are joking since the JVCs have none of the "funny" (yet very real) image issues.

These reports confirm my impression the Sonys' CAN yield a pastel look unless you are careful.

HERE's SOMETHING I WROTE A DECADE AGO ABOUT THE FIRST 1 CCD DV CAMCORDERS:

All NTSC video cameras have an Automatic Limiter Circuit (ALC) that keeps chroma levels from exceeding acceptable limits. (Red must be limited otherwise, when a TV separates luma from chroma, the red signal will contaminate the luma signal. The result is "cross color" artifacts that appear as "squiggles" on red objects.)

The ALC in a three-chip camera can analyze the red, blue, and green signals and more precisely alter the RGB signals before they are converted to luminance (Y) plus red (Y–R) and blue (Y–B) color difference signals.

The ALC in a single-chip camcorder can only alter the red (Y–R) and blue (Y–B) color difference signals. (Think of the ALC being nearly colorblind.) This can introduce unwanted color shifts. For example, if a very bright red flower is imaged, the ALC may slightly reduce the red difference signal relative to the blue difference signal causing the flower to become “pinkish.” (Because blue naturally has less luminance than red, it less likely to shift to magenta.)

-----------------

The A1 and HC1 are behaving like this, but they shoud NOT. There is no reason to limited red with HD and the CMOS feeds RGB to the DSP chip.

So I can't explain it -- except to say that FOR SOME REASON the Y-R component IS or BECOMES lower than the Y-B component. Either Sony missed this "bug" or it is inherent to the CMOS/DSP system they are using.

I wonder how Sony has been able to sell the A1 to pros? And, will they bring out a new A1 at NAB.

Or, is it a factory calibration issue that can be fixed?

[Dave Halliday]

Steve, I think you've described the problem very well.

[Graham Hickling]

Perhaps this is naive of me but .... if this isn't a problem in the A1, doesn't that suggest that it should be fixable via firmware in the HC1?