Expose to the right (ETTR) or underexpose?

[mjh]

Advertisement (gone after registration)

Does the analogue amplification (ISO setting) occur therefore within the sensor electronics prior to the signal being digitised and is the basic A-D conversion done by the sensor or the processor?

Basically there are three steps – after read-out, the voltage is amplified by a factor dependent on the selected ISO, then it is digitised and the digitised value may be amplified digitally (i.e. multiplied), also dependent on the ISO setting. Only the last step falls within the responsibility of the processor, a purely digital device.

 

For example, when the sensor’s base sensitivity is ISO 100 and the selected ISO is 3200, the analog amplification might be increased 32 times, or it might be increased 8 times, to be followed by a multiplication by 4, after the value was digitised. And so on.

 

I had assumed wrongly that this amplification was done by the processor and then the processor handled the A-D conversion. If it is being done at the sensor level, I accept that any result from the S2 with its CCD sensor would be wholly irrelevant to the M-240.

It depends … CCD sensors such as the sensors of the M9 or S2 (or S) are analogue devices and the ADC is an external circuit between the analogue sensor and the digital processor. With CMOS sensors the ADC circuitry is often integrated on the chip itself so the sensor output is digital – this is the case with the sensor of the new M.

[pelagia]

CCD sensors such as the sensors of the M9 or S2 (or S) are analogue devices ...

 

Wow, can we conclude that CCD is closer to film than CMOS?

What a nice speculation in favour of traditional leicaist.

[mjh]

Wow, can we conclude that CCD is closer to film than CMOS?

Many CMOS sensors are purely analogue devices, too, and those that aren’t are an analogue sensor and some ADC circuitry integrated on the same chip.

[algrove]

According to DXO Mark, the M9 applies NR to raws at all ISOs from 320 and up, and "This noise filtering is very slight but constant along the whole dynamic. It increases slightly with ISO."

 

On all of my M9's at ISO 160 they apply NR to raws. Take a 30 second exposure on your M9 and watch the 30 seconds of NR occur after releasing the bulb.

[mjh]

Take a 30 second exposure on your M9 and watch the 30 seconds of NR occur after releasing the bulb.

This dark-frame subtraction for eliminating hot pixels in long exposures has little to with ordinary noise reduction.

[douglasf13]

I'm sorry too, for not being able to make my point clear with this example. In your last post there are too many points raised outside the main subject of this thread about photography to be able to react to them in this thread without breaking up the main line of the conversation, so I leave it like it is for now.

 

Yeah, I think part of the problem is that this thread is being broken up into different topics:

 

- The M9's ISO doesn't need to be changed from base ISO, in terms of IQ, although this may not be practical in usage. This applies to the M9, but the technique doesn't apply to all cameras, like the D3.

 

- ETTR is certainly better in terms of SNR at base ISO, but it is unnecessary at ISOs higher than base ISO, because you could just lower the ISO and shoot with a "correct" histogram using the same exposure, rather than shooting at a higher ISO with a histogram to the right and then bringing the exposure slider down in post. This applies to all cameras.

[Jaybob]

Advertisement (gone after registration)

In this Forum we have seen recently interesting discussions about metering and ETTR. As I understand it would be preferable to ETTR combined with checking for possible clippings in the histogram. However in Dpreview's "Leica Talk Forum" Guy Plat explains on Aug 9 that he prefers to underexpose to get more dramatic images with punchy colors. Guy underexposes his images by putting camera settings at EV -2/3.

I can't recall where, but I also read that ETTR is not advisable: one should aim at a histogram with most of the tones in the middle of the graph.

What is in your experience and opinion in general the best way to expose?

 

Yeah, I think part of the problem is that this thread is being broken up into different topics:

 

- The M9's ISO doesn't need to be changed from base ISO, in terms of IQ, although this may not be practical in usage. This applies to the M9, but the technique doesn't apply to all cameras, like the D3.

 

- ETTR is certainly better in terms of SNR at base ISO, but it is unnecessary at ISOs higher than base ISO, because you could just lower the ISO and shoot with a "correct" histogram using the same exposure, rather than shooting at a higher ISO with a histogram to the right and then bringing the exposure slider down in post. This applies to all cameras.

 

1st and 266th posts...this has veered a tad off topic.

Edited October 29, 2012 by Jaybob

[douglasf13]

1st and 266th posts...this has veered a tad off topic.

 

Yeah, no kidding. It's been so long since I read the first post that I forgot what the main question was, which focuses on color.

 

Sandy's posts throughout the thread talk about how midtone color can be sacrificed with ETTR, so it's a trade off between slightly better SNR vs. slightly better color.

[pgk]

Sandy's posts throughout the thread talk about how midtone color can be sacrificed with ETTR, so it's a trade off between slightly better SNR vs. slightly better color.

Which is easy enough to prove to oneself with a little bit of experimenting:).

[tardegard]

On 8/14/2011 at 6:42 AM, denoir said:

Whether ETTR is a good idea or not depends on the sensor in question.

 

For the M9 the answer is that it is not. On the contrary, the M9 has far more latitude in the shadows than in the highlights. When in doubt it's much better to underexpose than overexpose. The highlights latitude of the M9 is extremely limited and overexposing with as little as a stop can result in blown highlights that you can't recover. Shadows are usually no problem and you can typically recover up to 4 stops with the only cost being added noise.

Exactly. 

[jaapv]

On 8/30/2020 at 2:28 PM, tardegard said:

Exactly. 

Errm..You are reacting to a seven year old post...

[tardegard]

8 hours ago, jaapv said:

Errm..You are reacting to a seven year old post...

Yeah but it was so interesting that I couldn’t resist... 😂

8 hours ago, jaapv said:

Errm..You are reacting to a seven year old post...