Best Way of Taking HDRs without Auto-Bracketing Feature

[Ysarex]

.........Look what happens when you increase the bit depth of the camera's ADC:

8 bit = 256 discreet tonal levels
10 bit = 1024 discreet tonal levels
12 bit = 4096 discreet tonal levels
14 bit = 16384 discreet tonal levels
..................

But how does this affect the ability to capture a wide dynamic range?

Very good question and admittedly I didn't really go there because I'd have to type a lot more. Theoretically each additional jump in bit depth adds a stop of DR. But just because we're adding the ability to record more data doesn't mean we're necessarily expanding (linearly) the outer limits of the data range. Engineering efficiency and design choices play a big role.

The OP is using an APS class camera so here's a look comparing the performance of a 10 bit ADC camera from some years back with Nikon's new D500 equipped with a 14 bit ADC -- thank you Bill Claff: Photographic Dynamic Range versus ISO Setting

The newer camera has picked up the ability to record 3 additional stops of dynamic range which fits the theory but I'm sure we could find example cameras that show some variation one way or the other.

Joe

 

[480sparky]

.........Look what happens when you increase the bit depth of the camera's ADC:

8 bit = 256 discreet tonal levels
10 bit = 1024 discreet tonal levels
12 bit = 4096 discreet tonal levels
14 bit = 16384 discreet tonal levels
..................

But how does this affect the ability to capture a wide dynamic range?

Very good question and admittedly I didn't really go there because I'd have to type a lot more. Theoretically each additional jump in bit depth adds a stop of DR. But just because we're adding the ability to record more data doesn't mean we're necessarily expanding (linearly) the outer limits of the data range. Engineering efficiency and design choices play a big role.

The OP is using an APS class camera so here's a look comparing the performance of a 10 bit ADC camera from some years back with Nikon's new D500 equipped with a 14 bit ADC -- thank you Bill Claff: Photographic Dynamic Range versus ISO Setting

The newer camera has picked up the ability to record 3 additional stops of dynamic range which fits the theory but I'm sure we could find example cameras that show some variation one way or the other.

Joe

I guess I'm not seeing ability to increase the dynamic range. Let's say a given sensor has a 10-stop DR range. And let's assume that's shooting 8-bit JPEGs.

So by "theoretically each additional jump in bit depth adds a stop of DR", shooting at 10-bit depth means you now have 12 stops of DR, 12 bits means 14 stops......... and when shooting at 14-bits means you get 16 stops of DR?

I can see increasing the tonal levels between stops, but I don't see shooting with an increased bit depth increases DR. 10 stops or 20 stops of DR, the bit depth merely increases the tone levels between the stops the sensor is capable of recoreding.

Kind of like there's only 12 inches in a foot. You can measure it out by 12 inches, 24 half-inches, 48 quarter-inches, 96 sixteenth-inches...... but it still all just one foot.

 

[Ysarex]

.........Look what happens when you increase the bit depth of the camera's ADC:

8 bit = 256 discreet tonal levels
10 bit = 1024 discreet tonal levels
12 bit = 4096 discreet tonal levels
14 bit = 16384 discreet tonal levels
..................

But how does this affect the ability to capture a wide dynamic range?

Very good question and admittedly I didn't really go there because I'd have to type a lot more. Theoretically each additional jump in bit depth adds a stop of DR. But just because we're adding the ability to record more data doesn't mean we're necessarily expanding (linearly) the outer limits of the data range. Engineering efficiency and design choices play a big role.

The OP is using an APS class camera so here's a look comparing the performance of a 10 bit ADC camera from some years back with Nikon's new D500 equipped with a 14 bit ADC -- thank you Bill Claff: Photographic Dynamic Range versus ISO Setting

The newer camera has picked up the ability to record 3 additional stops of dynamic range which fits the theory but I'm sure we could find example cameras that show some variation one way or the other.

Joe

I guess I'm not seeing ability to increase the dynamic range. Let's say a given sensor has a 10-stop DR range. And let's assume that's shooting 8-bit JPEGs.

So by "theoretically each additional jump in bit depth adds a stop of DR", shooting at 10-bit depth means you now have 12 stops of DR, 12 bits means 14 stops......... and when shooting at 14-bits means you get 16 stops of DR?

I can see increasing the tonal levels between stops, but I don't see shooting with an increased bit depth increases DR. 10 stops or 20 stops of DR, the bit depth merely increases the tone levels between the stops the sensor is capable of recoreding.

Kind of like there's only 12 inches in a foot. You can measure it out by 12 inches, 24 half-inches, 48 quarter-inches, 96 sixteenth-inches...... but it still all just one foot.

Sorry I ducked out -- grocery store.

The bit depth of the ADC establishes the theoretical limit of the system's DR. It doesn't guarantee it. By increasing the bit depth and as you note increasing the tone levels between stops we get cleaner, less noisy low-end stops that resist posterization; that data becomes more usable -- DR increase. Here's a chart and a few paragraphs from Cambridge in Colour.

It's up to the camera/sensor engineers to make the call in the sensor design whether the DR will in fact be increased or simply the tone levels will be smoother without a DR increase. So it's not a direct rule that more bit depth = more DR, but rather it's a general tendency for increased bit depth cameras to provide greater DR. The manufacturers have moved in that direction. Consider some Nikon history in this chart:

Photographic Dynamic Range versus ISO Setting

The D200 used a 10 megapixel Sony sensor that I'm pretty sure had a 10 bit ADC = 8 stops DR.
The D300 had a 14 bit ADC but didn't jump 2 stops, just 1 stop = 9 stops DR.
The OP's camera D3200 has a 12 bit ADC and one stop more DR than the D300 = 10 stops DR.
The D500 has a 14 bit ADC and a more efficient dual gain Sony sensor and = 10.5 stops DR.

Joe

 

[weepete]

To answer the OP, the easiest way to do it IMO is with shutter speed. So set yourself up with the camera on a tripod, spot meter on the brightest highlights and set the exposure when they meter 1.5 stops overexposed in the cameras meter. Then adjust the shutter speed to shoot at 2 stop intervals until the left edge of the histogram is inside it by about 1/4 to 1/3. Last step is to combine the multiple images in post.

 

[TCampbell]

Having a feature built into the camera is, in my opinion, a bit of a gimmick so that the camera maker can put another check-box on the list of features the camera includes. Photographers were taking HDR long before the cameras had an "HDR" mode.

You'll take bracketed exposures by selecting your fixed ISO and aperture values (which you wont change during the series of shots) and then you'll change only the shutter speed. If you alter the aperture, then you alter the depth of field (so focus across the image wont be consistent from frame to frame and you need consistent focus to have good stacking). By altering only the shutter speed you should be able to get frames which are consistent in every way except the amount of light gathered.

There are some videos (youtube is a great resource) that will walk you through the entire process.

Put your camera into spot-metering mode and take a meter reading of the element in your scene that you believe is the brightest and also meter the element in your scene that you think is the darkest. That establishes your exposure range to identify the shortest and longest shutter speeds that you should need. For most cameras using RAW mode, you don't need to shoot a frame in 1 stop increments... usually a 3 stop increment is sufficient but test it and back it down to 2 stop increments if you are unhappy with the results.

Put the camera on a tripod and take all the exposures from one end of the range to the other.... capturing all of this in RAW mode (not JPEG).

Feed the results into your image processing software. Several applications can do this including Lightroom, Photoshop, and others... but Photomatix is very popular for HDR -- it's purpose-built specifically for dealing with HDR.

 

[Ben1989]

I really really appreciate everybody's help on this. I went to Liverpool Cathedral over the weekend and taking all of the advice on board (sorry Tim I read yours too late). I'll post the picture below. I used Nik Efex Pro 2 to compile the HDR

 

[Ben1989]

 

[The Barbarian]

HDR is a technique, or rather a suite of different techniques. There are no HDR Police out there, although I've met a few who deputized themselves for that purpose. Do what makes good images, and let technique enforcers say what they will.

 

[droaingsong]

To answer the OP, the easiest way to do it IMO is with shutter speed. So set yourself up with the camera on a tripod, spot meter on the brightest highlights and set the exposure when they meter 1.5 stops overexposed in the cameras meter. Then adjust the shutter speed to shoot at 2 stop intervals until the left edge of the histogram is inside it by about 1/4 to 1/3. Last step is to combine the multiple images in post.

I am going to try this way too. Sounds like a lot of work but I like to see and compare the results.