RAW, ISO 1600

[fotobram]

I posted this as a reply to this one:

http://www.photocamel.com/index.php/topic,1220.0.html

Since I am rather curious to know the answer, I repeat this as a main topic;

As far as I know, it does not make sense to use ISO 1600 when shooting in RAW. The boost above ISO 400 is just software anyway, so can be done in postprocessing as well. Anybody has prove that I am wrong

As far as I know ( do not know where I did read it) it has the same effect as shooting at ISO 400 with 2 stops underexposure and subsequent raw development with 2 stops overexposure. That is probably why ISO 800 and 1600 are just option in the E300. Most usable for jpeg shooters.

[fotobram]

I like experimentation, but I am rather lazy when it comes to testing hard' or software.
Pavel, you can put this anywhere you like, as far as I am concerned, but I am really only interested in the E300 for practical reason.
I am so narrow minded

[Paul Shields]

It would be good to see examples. I'll try and come up with some if I have the time. The ISO boost does seem to be a software driven thing - to be honest I try and stick to 100 if at all possible, I'd use 50 if it was available.

[fotobram]

Yes, the burden of proof may be on me, but as said I am lazy and I do not want to re-invent the wheel if someone might already have done this. From earlier discussions I do have some indications, that what I suspect is true. Sofar I simply did not go any higher then ISO 400 and even that is too noisy for me most of the time. Most of the high ISO Oly pictures that I have seen sofar, I do not like that much. Sometimes I treat ISO 200 with neat image. Also in the old days I never liked the "grain effect" effects in film. I have never been very enthoudiastic about "artistic use" of that effect either.

What I have seen sofar is that Canon is the real winner when it comes to high ISO shooting.
We cannot be a winner in every aspect We do have to make choices.

[Raist3d]

Originally Posted by fotobram

Yes, the burden of proof may be on me, but as said I am lazy and I do not want to re-invent the wheel if someone might already have done this. From earlier discussions I do have some indications, that what I suspect is true. Sofar I simply did not go any higher then ISO 400 and even that is too noisy for me most of the time. Most of the high ISO Oly pictures that I have seen sofar, I do not like that much. Sometimes I treat ISO 200 with neat image. Also in the old days I never liked the "grain effect" effects in film. I have never been very enthoudiastic about "artistic use" of that effect either.

What I have seen sofar is that Canon is the real winner when it comes to high ISO shooting.
We cannot be a winner in every aspect We do have to make choices.

I want to kindly remind that personally I think this is a bit irrelevant because what matters is - if as a photographer there's a benefit while using the camera. I claim that regardless of the method there are two benefits over shooting JPEGS at ISO 800 & ISO 1600:

- The camera mettering will respond and measure light accordingly, as if it was using ISO 800 and ISO 1600. This is very important, particularly when you are using manual controls. It's just easier. You also get to verify the exposure of your shot if you want right there using the camera's LCD.

Proof: Set your evolt to spot metering and Program-autoexposure. Locate a point and use the exposure compensation until you get get a "perfect" exposure readout. Now without changing composition, spot where you are metering change it to ISO 800. Notice how the shutter speed decreases (i.e. faster shutter) and/or the aperture closes down.

- Particularly on the evolt, there's a difference in detail between RAW and JPEG. So it doesn't matter what method is Olympus using, you still get more detail at ISO 800 & 1600 in raw than jpeg

- Oh and I forgot, you may get more dynamic range, as RAW files carry more.

- Raist

[fotobram]

I agree with you about the advantage of RAW over JPEG. I never use JPEG. What I am wondering is, does it make sense to use RAW* ISO 1600 instead of -2 stops underexposure. I do not see your point about metering. When you use the method you describe above and set the +/- to -1 or -2 instead of setting ISO to 800 or 1600, you will also see the shuttertime and or aperture change.If you are talking of setting ISO higher on top of an underexposure compensation , then you are shifting to even higher ISO. ISO 1600 with -2 and afterwards +2 in RAW development would be ISO 6400 really. I have seen a series made in this way and they really look terrible.
About the LCD histogram, this could be the same trick the E300 sofware* is playing as that the Oly raw development sofware is said to do (see earlier comments).
I would be glad to assume that I am wrong in this all. Sofar I have seen nothing that really convinced me.
I admit I am not an easy guy to convince* It is a professional deformation grown out of 30 years of R&D experience (chemistry)
If you have convinced yourself in this, just ignore me. I do not want to irritate anybody with my bad mental health*

The key question in my mind* is, does the E300 apply extra signal amplification (with all the benefits and disadvantages) above ISO 400 or does it just mimick this via software ? Same for applying more gain to the sensor

As I understand no ISO specs are published by Kodak of the KAF-8300CE sensor. They just talk about quantum efficiency:
http://www.kodak.com/global/plugins/...CELongSpec.pdf

However I did read this:
"Total CCD noise is 16 electrons, which means that the dynamic range is 1600:1 - over 10 stops (64 dB)" , for what it is worth

[Raist3d]

Originally Posted by fotobram

I agree with you about the advantage of RAW over JPEG. I never use JPEG. What I am wondering is, does it make sense to use RAW ISO 1600 instead of -2 stops underexposure. I do not see your point about metering. When you use the method you describe above and set the +/- to -1 or -2 instead of setting ISO to 800 or 1600, you will also see the shuttertime and or aperture change.If you are talking of setting ISO higher on top of an underexposure compensation , then you are shifting to even higher ISO. ISO 1600 with -2 and afterwards +2 in RAW development would be ISO 6400 really. I have seen a series made in this way and they really look terrible.

First of all, I am not setting the exposure compensation to -2. The reason I mention the mettering advantage is that instead of worrying about setting the exposure to -2, I can "work within the frame of mind of ISO 1600" and then don't have to worry in post processing about cranking the exposure back. It's simply easier.

Also if I am shooting in full manual controls, it's much easier just to think of it as ISO 1600 instead of adjusting my shutter time to think about doing the -2 EV and then remember I have to do +2 EV back up. In other words- why worry about an extra detail while shooting and while postprocessing when I could have done it all by simply dialing ISO 1600? I don't understand what is the advantage to do otherwise.

And again, you can preview your image as if it was exposed to ISO 1600 on the LCD right there- yet another advantage.

About the LCD histogram, this could be the same trick the E300 sofware is playing as that the Oly raw development sofware is said to do (see earlier comments).

Ok, let's say it's a trick- so? Why do I care that really the file is ISO 400 if the LCD is giving me what the exposure would be at ISO 1600? I don't understand why this makes ISO 1600 shooting non desirable.

I would be glad to assume that I am wrong in this all. Sofar I have seen nothing that really convinced me.

I am not saying you may be wrong about the ISO 400 limit and the camera doing the 800 & 1600 for you. What I am saying is why does this negate in any shape or form the benefits of dialing ISO 1600 RAW vs having to worry about what Olympus did or didn't do, vs the method of underexpose + post process?

I admit I am not an easy guy to convince It is a professional deformation grown out of 30 years of R&D experience (chemistry)
If you have convinced yourself in this, just ignore me. I do not want to irritate anybody with my bad mental health

Well I am a computer software guy and I don't have R&D experiencel, and an aspiring photographer (I hope! ). I haven't convinced myself of what Olympus does- what I believe is that I don't understand why you say it's better to do the underexposure trick vs dialing ISO 1600.

The key question in my mind is, does the E300 apply extra signal amplification (with all the benefits and disadvantages) above ISO 400 or does it just mimick this via software ? Same for applying more gain to the sensor

I don't know and I wouldn't be surprised you are right about the ISO 400 limit. All I am saying is that it doesn't negate the benefit & convenience in shooting at ISO 1600 even if it's done by software on the camera.

As I understand no ISO specs are published by Kodak of the KAF-8300CE sensor. They just talk about quantum efficiency:
http://www.kodak.com/global/plugins/...CELongSpec.pdf

However I did read this:
"Total CCD noise is 16 electrons, which means that the dynamic range is 1600:1 - over 10 stops (64 dB)" , for what it is worth

If what you really want to find out is what Olympus does, that I don't know, and honestly I don't care. I see the consequences already of a camera sensor being pushed beyond its limits (i.e. a lot of noise at ISO 1600), so software or hardware the price is the same. But the thread started with the value of shooting at ISO 1600 in RAW vs jpeg, or why even shoot ISO 1600 anyway - to which I say there's several advantages (mentioned above).

So this is what I am disputing:

As far as I know, it does not make sense to use ISO 1600 when shooting in RAW. The boost above ISO 400 is just software anyway, so can be done in postprocessing as well. Anybody has prove that I am wrong Huh Huh

- makes sense because you don't have to worry about remembering to under expose and you dont' have to post process for that

As far as I know ( do not know where I did read it) it has the same effect as shooting at ISO 400 with 2 stops underexposure and subsequent raw development with 2 stops overexposure. That is probably why ISO 800 and 1600 are just option in the E300. Most usable for jpeg shooters.

- makes sense to shoot RAW because you still get more details than shooting ISO 1600 jpeg on the evolt.

So to finalize the point: why would anyone in their right mind shoot at ISO 400 and worry about underexposing, to then worry about overexposing in post process, and not have the advantage of seeing your shots int he LCD of the camera in the field along with histogram exposure information to help you make decisions on reshooting?



- Raist

[fotobram]

I think you are right. By no means I had the intention of changing the habits that work fine for you.
Maybe my initial statements were a bit provocative. My only motivation was and is to find out what happens in the E300.
For me it does not make a difference if I have to think in ISO 1600 or ISO 400 -2 +2
Looking at a bright postview (you are talking about preview ?) is an advantage indeed. However a histogram that might be a fake one , just puts me on the wrong leg (is that an English expression, or just a badly translated Dutch one ?).
I just want to find out if E300 - right of the (real) histogram shooting has the advantages that Pavel is talking about in a performance sense.

[fotobram]

Originally Posted by th0t

Hi fotobram,

Keep posting. Just because the idea behind this thread may be controversial shouldn't deter you from continuing to question things in the future. Who knows? Next time you might spark some interesting debate and learning (on our part). We all need to learn to think "outside the box" more frequently. I like the statement in Pavel's signature that goes: "To find the answers - question them!"

Sounds like you consider my questions answered. I don't*

[Paul Shields]

Originally Posted by Pavel

So - this is all in a welcoming spirit - while still disagreeing.

Pavel - you should be a politician

[bavanor]

Ok, this thread has got me wondering, and upon wondering I remebered these pictures I took in savanah. They are of an old telegraph police street box. I shot them in iso 800, 1600 and 3200 with the E-1. The iso 800 is taken @ 56mm, f2.8, and 1/80 sec. The iso 1600 is taken @ 60mm, f2.8 and 1/60 sec. The iso 3200 photo is taken @ 60mm f3.5 and 1/80 sec. Now I processed them in Raw Shooter essentials with out changing anything (just matched up the white balance). Then in photoshop I sharpened them all the same and shrunk them to post on the web.

Check out the results.

Aaron

[bavanor]

and here is the iso3200 photo

Aaron

[fotobram]

Pavel,

I think the tone has nothing to do with this. It does not change facts nor understanding.
As I have said before, I think I* understand your theories and I accept they work fine for Canon, Nikon or whatever.
I have seen no proof that they work for the E300 and "someone at DPREVIEW" (he is not alone) might be right.
It looks like I have to find out for myself. No problem.
That is the situation for me at the moment
Maybe I can find the time and the patience sometime to find out or maybe I will stick to ISO 400 max. as I did sofar.
After all the E300 is not brilliant as it comes to high ISO anyway.
Or maybe I will buy a Canon someday.*
Thanks for you attention. You might not believe, but I have learnt a few things

[fotobram]

Thanks Pavel,

I appreciate that you put so much effort in this.
Again said that I think I understand this all and try to remember it while shooting.
Most of the time I forget all about things like this while I am actually shooting.
I am more going on instincts then.
In fact I have noticed this already in practise with 100,200,400 ISO shots.
As I have said, sofar I did not go any higher.
I am really not arguing that we should not apply this above ISO 400.
My question is more, are we ? Or are we just thinking we can apply this when we shoot over ISO 400 on a E300 ?
What use is a histogram, that is not the real one ?
Maybe my mind is on the wrong track. I can accept that too.
Please stop explaining now, because I am claiming too much of your time

BTW I have noticed that all real bright RAW's on the E300 show kind of purple in the highlights when tuned down too darker developments.

[Raist3d]

Originally Posted by fotobram

Pavel,

I think the tone has nothing to do with this. It does not change facts nor understanding.
As I have said before, I think I understand your theories and I accept they work fine for Canon, Nikon or whatever.
I have seen no proof that they work for the E300 and "someone at DPREVIEW" (he is not alone) might be right.
It looks like I have to find out for myself. No problem.
That is the situation for me at the moment
Maybe I can find the time and the patience sometime to find out or maybe I will stick to ISO 400 max. as I did sofar.
After all the E300 is not brilliant as it comes to high ISO anyway.
Or maybe I will buy a Canon someday.
Thanks for you attention. You might not believe, but I have learnt a few things

What I am truly curious about is what is the advantage you see in using ISO 400 and doing the underexposure trick vs doing a plain ISO 1600 setting on the evolt. If the result is going to be the same, why not choose the path that leads to fewer tracking of details, less opportunity for error and less post processing? Is there an advantage in doing the trick?

[fotobram]

Originally Posted by Raist3d

What I am truly curious about is what is the advantage you see in using ISO 400 and doing the underexposure trick vs doing a plain ISO 1600 setting on the evolt. If the result is going to be the same, why not choose the path that leads to fewer tracking of details, less opportunity for error and less post processing? Is there an advantage in doing the trick?

No, I see no advantage. Your method has at least some practical advantage. Just wondering about the technical advantage.

I just did a few quick test shots with my E300, and I must say I just do not like the results above ISO 400
Because I consider the results poor anyway I cannot decide if there is a real difference in result between the two methods described.