[it_inspector]

Originally Posted by clevermax

I am surprised to see this statement. RAW is what the sensor actually sees (data straight from the sensor) and the WB settings doesn't change the actual RAW image data. The WB setting value will just get embedded in the RAW files as a metadata for the RAW application to interpret and display the RAW file accordingly. Even the camera's LCD displays RAW according to the WB setting under which it is shot but that doesn't mean the image data has been manipulated according to the WB setting. All this as per my limited knowledge and understanding.

Basically, RAW is as is. Either the camera or the computer reads the raw image, then references the settings like WB, sharpness, saturation, etc. to determine how to display the image. Any of those settings can be changed later, because it's only a guideline, not set in the RAW data itself.

When it comes to Jpeg, these settnigs are 'hard coded' in the actual pixel data - so it's not possible or difficult to change them later.

In camera, the WB settings are taken into account when the Jpeg is made from the RAW data captured.

Experts, correct me if I am wrong.

Originally Posted by tsk1979

I don't know what to say. I suggest you look up the camera supplied software manual for your camera, and look up how to select WB in RAW.
In a RAW image you can select any WB from around 1600 to 10,000. RAW is basically RAW sensor data, there is no concept of WB in RAW. Maybe you are shooting RAW+JPEG, and then looking at JPEG? RAW files in canon have a .CR2 extension.

Originally Posted by SPARKled

@it_Inspector: You are using the Nikon D3x and the D3 right? So you must be having Capture NX for sure as your RAW convertor. The WB can easily be fixed using Capture/ Capture NX. Infact one of the major reasons of shooting RAW for me is WB adjustment in post processing.

I know what RAW image file is, I goto court to invalidate RAW image files in front of jury and judges. So i do know a tiny bit about image formats

Yes WB can be corrected during PP when original RAW data is available, but it is not accurate since there is no REAL LIFE calibration available.

Correct me if i am wrong but as far as i know, only available options available in PP are Auto, Daylight, Cloudy, Shade, Tungsten, Fluorescent, Flash Custom, As shot, Custom. Custom being usually what was CALIBRATED on the scene or AS SHOT.

Simple experiment anyone can do at home. Grab some transperency sheets, baking paper and a table lamp. Print transperency sheets with different color shades (blue, green, red) as if its lighting in a concert. Then put the transperency sheet in front of the table lamp and point it towards the subject. Take the picture in AUTO with RAW, AUTO with Custom White balance with RAW, AUTO with Custom White balance with Jpeg. Then try correcting all of the files in Photoshop/Lightroom/Camera Raw/Capture Nx.

Think of Custom White Balance as Color Correction for Sensor.

Originally Posted by tsk1979

All cams will have hot pixels(speaking from experience) esp if you shoot with dark frame noise reduction switched off.
So how long were the exposures, and was dark frame subtraction off or on in the camera?
Moreover, to compare apples to apples, both cameras have to be at same ambient temp. Even a 5 degree difference in temp, and few minute of difference in exposure time can result in a huge difference

You are confusing Noise with Hot pixels. Like i said earlier Noise comes down and eventually fades away as sensor cools down. I am talking about Hot Pixels.

Defective pixel - Wikipedia, the free encyclopedia

Hot Pixel is effectively a Defect in the sensor. As comparing Apples to Apples, sensors survive in Arctic and in 50 degree heat in Centre Australia and i have seen 1hr exposure with D3 but that has nothing to do with Hot Pixels.

Originally Posted by shajufx

Bouncing light is something which I really want to experiment and my cam is D60, so sticking to nikon itself would give me some advantage when it comes to different in-built settings. So far no idea if other brands can actually function well on nikon.

Yes I heard a lot about Nikon CLS and D60 supports it. My interest is to the tilting 90° up, 180° to the left and 90° to the right. Now the pricing part, is there any bargaining possible for a 1 month used flash ? I know lenses do not lose much value as seconds, but what about the flash ?

One cheap but very useful accessory for flash is PC sync cable or flash mount extension cable. Just make sure its as long as you can get, it gives so much creative flexibility.

Also Flashes loose value as time passes, just like bodies.

[tsk1979]

There are 2 more options
1. Color temperature - Here you can set any color temperature you want
2. Color picker - Here you click on the point which you think should be grey/white. For example, lets say you shoot a room and there are some white papers lying there. As shot, those papers appear yellow. So all you do is click on the white papers, and you will have instant color correction. This is crudely similar to caliberation.

Now regarding Hot Pixel,

From this webpage, and many others
The Imaging Resource Hot Pixels Article

In addition to dark current, temperature is also a factor in creating hot pixels. The higher the temperature, the higher the charge leakage. A 10 degree change in temperature can noticeably change what the CCD reports.

So for short exposures, you will not see hot pixels. If temperature or exposure time increases number hot hot pixels will also increase due to charge leakage.
Another link from Ken Rockwell
Hot Pixels

So if you are saying that the number of hot pixels will remain same of temperature ranges and exposure times, you are wrong.

Moreover the wikipedia link you posted is for LCD Hot Pixels. I am talking about CMOS and CCD sensors, and over here there are 2 types of permanent dead pixels
1. Dead pixel(always black)
2. Stuck pixel(always white)

Anyways dpreview does a 30 second exposure test for Hot Pixels for almost all cams

[clevermax]

Originally Posted by it_inspector

Yes WB can be corrected during PP when original RAW data is available, but it is not accurate since there is no REAL LIFE calibration available.

So you meant to say you always calibrate WB on the frame you want to shoot rather than using preset values or custom values available in the cam? Then it is going to be a tedious task shooting each frame.

Anyway, that's deviating from the point. The point is, RAW image file can be corrected for WB to the desired/optimum look during PP - no matter what kind of WB setting you had in camera while shooting RAW.

Originally Posted by it_inspector

Correct me if i am wrong but as far as i know, only available options available in PP are Auto, Daylight, Cloudy, Shade, Tungsten, Fluorescent, Flash Custom, As shot, Custom. Custom being usually what was CALIBRATED on the scene or AS SHOT.

Adobe Camera RAW allows one to change it smoothly by selecting the correct values for hue and tint - they are sliders and not preset values.

Originally Posted by it_inspector

Simple experiment anyone can do at home. Grab some transperency sheets, baking paper and a table lamp. Print transperency sheets with different color shades (blue, green, red) as if its lighting in a concert. Then put the transperency sheet in front of the table lamp and point it towards the subject. Take the picture in AUTO with RAW, AUTO with Custom White balance with RAW, AUTO with Custom White balance with Jpeg. Then try correcting all of the files in Photoshop/Lightroom/Camera Raw/Capture Nx.

All those RAW shots will have the same RAW data except for minor differences that can occur in the lighting of the subject while you shoot different frames over time.

For example if you shoot the exact same frame with the exact same lighting

using
RAW & Tungsten WB and
RAW & Daylight WB

The raw image data for both images will be the same but the WB setting which is stored as metadata in the RAW file will be different and because of that, the pictures will look different (in terms of WB) in camera's LCD or in the RAW editing application while you open it. But both the images can be edited to achieve the exact same WB look while you make a jpeg out of them by changing the WB values in Adobe Camera RAW.

We're not talking about shooting jpegs as it is different - for jpegs the WB setting at the time of shooting matters.

Originally Posted by it_inspector

You are confusing Noise with Hot pixels. Like i said earlier Noise comes down and eventually fades away as sensor cools down. I am talking about Hot Pixels.
Defective pixel - Wikipedia, the free encyclopedia

Surprised again. I thought we were talking about hot pixels in sensor - the link is actually for defective pixels in LCD screen.

Originally Posted by tsk1979

So if you are saying that the number of hot pixels will remain same of temperature ranges and exposure times, you are wrong.

True, some hot pixels will show only after a particular time of exposure, and it can vary from one hot pixel to another, AFAIK

[it_inspector]

Originally Posted by tsk1979

There are 2 more options
1. Color temperature - Here you can set any color temperature you want
2. Color picker - Here you click on the point which you think should be grey/white. For example, lets say you shoot a room and there are some white papers lying there. As shot, those papers appear yellow. So all you do is click on the white papers, and you will have instant color correction. This is crudely similar to caliberation.

Option is there and a crude workaround is there. Your decision what to choose. I only put it forward because Sony A550 had the issue with WB and someone asked for help on deciding whether to buy it or not. And that person is a beginner.

I can argue Calibration vs Crude method about how DSLR's get calibrated at the factory and usually don't need recalibration while Lcd's (even Eizo) needs calibrating every Week. But honestly final decision is yours. Trust uncalibrated Lcd or a Calibrated DSLR with on the spot calibration which costs nothing.

Originally Posted by tsk1979

Now regarding Hot Pixel,

From this webpage, and many others
The Imaging Resource Hot Pixels Article

So for short exposures, you will not see hot pixels. If temperature or exposure time increases number hot hot pixels will also increase due to charge leakage.
Another link from Ken Rockwell
Hot Pixels

Those 2 resources are one of the worse to follow. Kenrockwell himself makes it clear that he likes to make up things from thin air and has a brain of troubled 9yr old.

About Noise generated with long exposures its mostly Salt-and-pepper noise which can be correctly largely using Dark Frame Subtraction. Which is completely different to Hot Pixels.

Originally Posted by tsk1979

So if you are saying that the number of hot pixels will remain same of temperature ranges and exposure times, you are wrong.

WRONG. Hot Pixels show at All Shutter Speeds. Temperature on the other hand affects Hot Pixels more than anything else. Temperature change can cause Hot Pixels and it can also repair Hot Pixels.

Originally Posted by clevermax

So you meant to say you always calibrate WB on the frame you want to shoot rather than using preset values or custom values available in the cam? Then it is going to be a tedious task shooting each frame.

Its not really a tedious task to be honest. Just like changing lenses is a tedious task for a beginner it only takes getting used to. More over if i am doing a studio session, i only need to do Custom white Balance once which will take less than 30 seconds while correcting each file will be well even with batch process it will take fair while.

Also i do not do Custom White Balance each shot. It depends upon the need with varying lighting conditions. But i do it quite a bit. I would say 50%+ for personal shots while 100% for anything i am shooting for clients.

Originally Posted by clevermax

Anyway, that's deviating from the point. The point is, RAW image file can be corrected for WB to the desired/optimum look during PP - no matter what kind of WB setting you had in camera while shooting RAW.

Corrected - Sure
Optimum look - Not as easy

Originally Posted by clevermax

Adobe Camera RAW allows one to change it smoothly by selecting the correct values for hue and tint - they are sliders and not preset values.

Sure, if you believe your Eyes + uncalibrated/calibrated - Lcd/Crt/Projector is more efficient than Calibrated sensor in your DSLR.

Also Preset values are Calibrated in a Work Flow which is Calibrated from Sensor to Printer.

Originally Posted by clevermax

All those RAW shots will have the same RAW data except for minor differences that can occur in the lighting of the subject while you shoot different frames over time.

For example if you shoot the exact same frame with the exact same lighting

using
RAW & Tungsten WB and
RAW & Daylight WB

The raw image data for both images will be the same but the WB setting which is stored as metadata in the RAW file will be different and because of that, the pictures will look different (in terms of WB) in camera's LCD or in the RAW editing application while you open it. But both the images can be edited to achieve the exact same WB look while you make a jpeg out of them by changing the WB values in Adobe Camera RAW.

Have you tried doing it yourself ???

If yes then you are different to thousands of Concert Photographers who shoot Concerts, especially where different colored lights are used.

If not, then i seriously suggest you do it, since Concert Phtographers spend a fair bit of time setting Custom White balance.

Please do not confuse Concert Photographers with normal Photographers since above example is seen by Concert Photographers all the time.

There is a reason why i specifically suggested this experiment.

Originally Posted by clevermax

We're not talking about shooting jpegs as it is different - for jpegs the WB setting at the time of shooting matters.

I am also reffering to RAW files. BTW All the Crude methods suggested so far also work on Jpeg's altough not 100% effective.

Originally Posted by clevermax

Surprised again. I thought we were talking about hot pixels in sensor - the link is actually for defective pixels in LCD screen.

Similar defects can also occur on CCD or CMOS image sensors in digital cameras. In these devices, defective pixels fail to sense light levels correctly, whereas defective pixels in LCDs fail to reproduce light levels correctly.

Hot Pixel is a DEFECTIVE PIXEL. Hence i said each time to not confuse it with Noise. Noise starts increasing as sensor gets hot and fades awat as sensor cools down. Hot Pixel will stay all the time.

Hot Pixel will usually show first time with a long exposure but after that it will show on ALL Shutter Speeds and all Apertures and on every single shot.

Originally Posted by clevermax

True, some hot pixels will show only after a particular time of exposure, and it can vary from one hot pixel to another, AFAIK

WRONG. Hot Pixels show at All Shutter Speeds.


Since I know my message has still not gone across. I am attaching a picture with Hot Pixel. Its not my picture and is borrowed from a local Photographer.



Thanks

[Rudra Sen]

Originally Posted by it_inspector

Since I know my message has still not gone across. I am attaching a picture with Hot Pixel. Its not my picture and is borrowed from a local Photographer.

Can't see the picture here. Post it again.

[shajufx]

More than 70% of the discussion is way over my head !!! Do I need to worry so much about all these ? NO way ! But one thing I can boldy suggest: If you guys could post some examples on what you are discussing, even a newbie like me could catch a lot from here.

[it_inspector]

Originally Posted by Rudra Sen

Can't see the picture here. Post it again.

Here you go!

Originally Posted by shajufx

More than 70% of the discussion is way over my head !!! Do I need to worry so much about all these ? NO way ! But one thing I can boldy suggest: If you guys could post some examples on what you are discussing, even a newbie like me could catch a lot from here.

Sure will try to post more examples next time.

[tsk1979]

Originally Posted by it_inspector

You are confusing Noise with Hot pixels. Like i said earlier Noise comes down and eventually fades away as sensor cools down. I am talking about Hot Pixels.

Defective pixel - Wikipedia, the free encyclopedia

Hot Pixel is effectively a Defect in the sensor. As comparing Apples to Apples, sensors survive in Arctic and in 50 degree heat in Centre Australia and i have seen 1hr exposure with D3 but that has nothing to do with Hot Pixels.
.

Originally Posted by it_inspector

Those 2 resources are one of the worse to follow. Kenrockwell himself makes it clear that he likes to make up things from thin air and has a brain of troubled 9yr old.

About Noise generated with long exposures its mostly Salt-and-pepper noise which can be correctly largely using Dark Frame Subtraction. Which is completely different to Hot Pixels


WRONG. Hot Pixels show at All Shutter Speeds. Temperature on the other hand affects Hot Pixels more than anything else. Temperature change can cause Hot Pixels and it can also repair Hot Pixels.

Since I have taken shots with 350D ranging from few seconds to 30 minutes(without dark frame noise substraction) I have observed differently. For shorter exposures, there are no "hot pixels", but for longer exposures they are.


You are confusing Hot Pixels with Stuck/Dead pixels.

Regarding links, since the links I have posted are all wrong and the worse places for information, can you post some relevant links?
This Defective pixel - Wikipedia, the free encyclopedia is relevant for LCD TV, not for digital cameras.

Last but not the least, the pic you posted does not look like a Hot Pixel at all. Looks more like an artifact or stuck pixel.
Here is a 100% crop showing a "hot Pixel"


Now coming to your point of defective pixels, they are stuck pixels.
A stuck pixel will appear in all the shots. A hot pixel will appear only on the Longer exposures.
Take a 10 minute exposure in a dark room with lens cap on. then wait some time, and take a 10 second exposure with lens cap on in a dark room.
In the 10 minute exposure you will see a lot more hot pixels.
However, if your camera has dead pixels, they will show up in all the shots.

For the record, I did a test on my camera, and there were 2 or 3 "Stuck pixels" on my camera(a very very insignificantly small percentage of the 8 MP)

There are lots of web programs which can analyze the image from your cam, and tell you the exact number of Dead/Stuck and Hot pixels.

So By definition
Stuck pixel = a pixel that always reads high (maximum) on all exposures.
Hot pixel = a pixel that reads high on longer exposures.
Dead pixel = a pixel that reads zero (black) on all exposures.

A stuck pixel or a dead pixel is a defective pixel, but a hot pixel is not a defective pixel, and will show up only on very long exposures.
Typically many cams have 1 or 2 hot pixels at 30 seconds exposure, and as the exposure time increases, more and more pixels go hot.
Compact cams have a lot more than larger sensors.

Hot Pixel will usually show first time with a long exposure but after that it will show on ALL Shutter Speeds and all Apertures and on every single shot.

So if I get a hot pixel, after that it will show in all the pictures at all shutter speeds?
Well here is a long exposure I took long time back. This is equivalent of over an hour of exposure. Hot pixels have been circled. There are many because I did not take dark frame substraction, and ambient temp was very high.
For actual exact image please look at link at the end of this post



Now according to what you said, in all my subsequent shots, I should be seeing these "hot pixels"?
Well I do not

Since the JPEG compression on the site is killing the quality, here is the link to image on smugmug where you can see it more clearly
http://tanveer.smugmug.com/Travel/Hi...3_ApvjF-XL.jpg

[Rudra Sen]

Originally Posted by it_inspector

Even RAW can't do anything when White balance is off. And i only shoot RAW.

Why one would shoot with white balance off? I don’t get the reason! Is there any special advantage? If yes then I would like to know what is that. I haven’t shot anything till date without white balance on and it is at AUTO setting from the day one. I don’t think I’ve lost any or got any wrong colour information.

Originally Posted by it_inspector

Optimum look - Not as easy

Subjective statement. What is optimum is a question as always. You see something before shooting and want it to reproduce faithfully is something can be close to optimum. Or that’s what I believe. Can I enhance that? Sure I can if I have all the information in my RAW file. Do I need to do that? Yes and no. It depends greatly on final application.

Originally Posted by it_inspector

Hot Pixel is a DEFECTIVE PIXEL.

Do not buy this at all. If that’s the case then one should be seeing those defective pixels right through all images. That doesn’t happen.

Originally Posted by it_inspector

Hot Pixel will stay all the time.

Where? On the same place? If I may ask you from where are you getting all these information?
Three shots below were taken within a span of hour and half. I’ve marked the areas where hot pixels were sitting. All three were with roughly 30 seconds+- time.

Originally Posted by it_inspector

WRONG. Hot Pixels show at All Shutter Speeds.

Really? Then how come I don’t see them all the time? Sample as what you have put up is not hot pixel at all. What I see from TSK is more like it.

Originally Posted by it_inspector

Noise starts increasing as sensor gets hot and fades awat as sensor cools down. Hot Pixel will stay all the time.

Oh! Then how come I didn’t find any hot pixel on any of my May ’07 shooting at UP? I was shooting at 45 degree.

At the end of the day one likes to see the result. There are pros and cons with every single technology. There are ways to bypass those and produce good result. What concert photographers do is not important unless that is somebody’s bread and butter. Let us be real here. I’ve seen hot pixels from Nikon’s top of the line cameras. That doesn’t say anything. Similarly I’ve seen excellent result from 6000/- point and shoot camera.

it_inspector. My humble request to you: Be real. Searching internet for all the solutions and beating around the bush is the same in the sphere of digital photography. If there are white pixels/dead pixels/hot pixels.. Too bad. One must know how to work around that. And if you know that then please guide us.

I’m sorry if I’m sounding rude here. It’s not my usual self and trust me on this. I’m flabbergasted with your series of replies. There are many members here interested in photography. Some of them are advanced and some of them are just getting in. Your wealth of knowledge should be guiding them and not the other way round.

[jkdas]

! ! ! couldnt help!

Rudra sir, its all black.

[Rudra Sen]

What is black here JK?

[jkdas]

Originally Posted by tsk1979

Now according to what you said, in all my subsequent shots, I should be seeing these "hot pixels"?1
Well I do not

Since the JPEG compression on the site is killing the quality, here is the link to image on smugmug where you can see it more clearly
http://tanveer.smugmug.com/Travel/Hi...3_ApvjF-XL.jpg

Originally Posted by jkdas

! ! ! couldnt help!

Rudra sir, its all black.

Originally Posted by Rudra Sen

What is black here JK?

I couldnt see the difference when viewed in small screen and coz of the reason mentioned by tsk ( the link by tsk helped differentiate).

Some posts are too technical for me but I learned something today ( about white balance).

[clevermax]

This forum has always been helping me in learning things related to automobiles and I am sure some threads like this one and the image threads are helping many of us to learn and understand few things about photography as well - thanks to the experts who are always there to clear our doubts and misconceptions about many things.

I'd like to express my thanks again to the experts like Rudra & Tanveer who have actually helped many of us here not to develop misconceptions about few things which were being discussed here lately.

I am not an expert in this field, many of us are not. So its natural that many of us folks fail to distinguish between showoff and real expert talk and some seem to take advantage of this. But if one tries to learn things on his/her own and try to interpret what seems to be OHT talk (Over head transmission) we can all really understand what is real and what is not.

Cheers & keep clicking folks.

[Torqueguru]

@Rudra
@TSK
Thanks for the invaluable insight.
@clevermax
Well said.

Regards,
TG.

[shajufx]

After all that discussion, I am now searching for 'hot pixels' and 'dead pixels' in my pics, but not lucky so far.

I found some dirt marks on latest pics though (visible only when this specific area comes in light shades), did some image sensor cleaning as instructed in the manual, but not yet being successful in removing the mark. Checked dust on the lens too. I tried lifting the sensor and cleaning inside with an earbud. The mark has lightened but not fully gone. Here is a pic of the sky. I havent got a blower yet. Any suggestions ?