A900 findings

[douglasf13]

Andrey, I think it should be mentioned that: one, I assume you're speaking of spot metering, and not matrix. Two, sometimes not using ETTR means that highlights may be blown out. I've seen many misunderstand ETTR and not let any highlights blow at all, which results in the underexposure of midtones/subject. Your general advice of making sure that midtones are in the right place sounds good to me.

 

[hardloaf]

Bottom line on in camera cam meters almost every company Canon, Nikon and my bet Sony as well tune the meters to underexpose slightly to protect themselves from about a million daily complaints from blown highlights. Honestly this is just a fact of life and the same with using AA filters with complaints by the millions from Moire. Seriously and I don't mean too offend anyone but these cameras are designed with the amateurs in mind that just stick the camera on auto and go, plus jpeg all the way. That is the design mindset. Obviously they are made for Raw shooting and such as well and in more capable hands can adjust at will. But also every Raw converter will render each Raw differently as well and some will be more underexposed and some will not. The end of the day there is no standard and you will have to fine tune for your raw converter. My Phase back is about 1.25 stops under in LR and slightly under in Raw developer by about 1/2 stop. Also on meters what many folks don't even think about is where the weight area of the meter sits, so that throws another curve in the mix.

Gray point is the base point which light meter is calibrated to. There used to be an old film 18% gray standard (2.5 stops from saturation point at the top), than at some moment all vendors switched to 12.7% (about 3 stops). They all round up them differently, so there is some small variations, f.e. it can be 12.5%. Then they switched to even lower values. My A900 is about 10% gray (3.33 stops), but it looks like there is slight variations in different camera samples. This all is based on green channel only (not sure if they use same kind of green as on sensor though).

You right that this number regulates amount of captured highlights, but I don't think that it has anything to do with amateurs - all light meters have their gray point and that's how they work. I don't know exact reasoning behind 18% gray, but this is the area which humans like a lot and where we have highest sensitivity to details and colors. Where 12.7% comes from I can only guess, probably something with getting round power of 2 which is important for digital (it's very close to 2^3=8). Comes with a "bonus" of more highlights at cost of shadows, but requires compensation because 12.7% perceived as underexposed. Now recent crop of cameras like A900, D3X, 5D2 went even farther. A900 calibrated to 10% gray and picture exposed around this number will definitely look dark. Most likely vendors just optimized light meters to place gray point in the best spot of a sensor, which is a good thing - they can afford this because sensors are better and shadows are much cleaner now. Bad thing here is that pictures need to be compensated now. In RPP this means Compressed exposure correction about 0.7-0.8 for A900. It absolutely must be corrected!

Other converters do this behind a curtain and apply some exposure corrections automatically, usually badly and based on their own (mis)understanding, tone curve is getting involved so that's were all this confusion comes from. I completely disagree with this approach and in RPP you'll actually see picture exactly as it was captured and you need to define your correction specifically and make it default.

 

[douglasf13]

Andrey, if you have a scene with very low dynamic range that has no risk of clipping shadows and highlights, where do you place middle grey? -1, 0, or +1 EV? Which is the ideal setting for best color, noise, resolution, etc? I'm debating whether there is any reason for me to be using Zone -1. I've been metering middle grey at +0, or essentially +.5 with Zone turned off. Thanks.

 

[edwardkaraa]

I think the 18% came from slide film and color printing paper latitude of 5 f/stops. Slide film had usually a latitude of 2 1/3 over and 2 2/3 under. I also heard that Sekonic grey point is currently at 13% instead of 18%, probably to compensate for the increase in modern film latitude. I think Andrey is making a very valid point about meters. I tested my A900 meter against a Sekonic and the readings were practically similar (if anything, the A900 overexposed very slightly over the Sekonic by 1/10 to 1/20 f/stop). This means that while the meter reading is basically correct, the real iso sensitivity of the camera may not be exactly as advertized. DXOmark suggests about 1/2 stop below advertized numbers.

This discussion is getting more and more interesting

 

[Quentin_Bargate]

Fascinating, but 320 ISO best? I think not.:lecture:

Quentin

 

[jonoslack]

Fascinating, but 320 ISO best? I think not.:lecture:

Quentin

I think an awful lot of this discussion hinges around the definition of the word
best
Which may be different for different people (certainly is for me).
The 'no ETTR' idea is music to my soul, because I'm more interested in colour and an absence of blown highlights than I am in shadow noise.

It's all really interesting and constructive though

 

[douglasf13]

Agreed, Jono. Some may find the very small difference in noise between ISO 200 and 320 more important than destroyed shadow detail.

As far as ignoring ETTR, I think we need to be careful with this sentiment, as there are different definitions of this term, as well as the fact that some raw converters do a better job than others. Some view ETTR as exposing everything to the right without blowing any highlights, but that can be fundamentally flawed, because it forces some shooters to try to save insignificant highlights, like speculars, which can lead to underexposure of midtones. Overall, I think Andrey's recommendation to keep the subject/midtones in the Zone 4,5 or 6 range is a good way to look at things. Unfortunately, from what I've seen, matrix metering isn't a consistant way of assuring this.

 

[Eoin]

Fascinating, but 320 ISO best? I think not.:lecture:

Quentin

Hmmm!, I like shooting available light portraits, I also happen to think for my criteria 320 ISO offers the best compromise in these situations. I am too lazy to bother changing back down to a lower ISO because quite frankly I don't see that much difference in print, infact I quite prefer the slightly different structure it imparts to monochrome prints.

One mans medicine is another mans poison!.

 

[Terry]

Other converters do this behind a curtain and apply some exposure corrections automatically, usually badly and based on their own (mis)understanding, tone curve is getting involved so that's were all this confusion comes from. I completely disagree with this approach and in RPP you'll actually see picture exactly as it was captured and you need to define your correction specifically and make it default.

Which takes us back to the thread two weeks ago (when I got the camera) when we talked about setting camera defaults and in conjunction, zeroing out the Lightroom "behind the scenes" corrections.

 

[douglasf13]

Right on, although keep in mind that those LR settings are a rough estimate, and they won't be nearly as accurate as RPP in showing what your RAW really looks like.

 

[Georg Baumann]

Anyone around here had a chance to compare results of RPP and C1?

Andrey, I realise you are the manufacturer, however, I would think you did make some form of empirical studies comparing your product with the output of other converters, and it was that I was curious about.

In what fields did you find RPP provides an advantage over C1 for example.

I am not biased at all, I have not used C1 or RPP, I only used Sinar xposure and LR.

Having said that, I am now shooting with the alpha and of course I am interested in best possible file quality. I am also shooting with a phase one, did you have a chance to try your RPP with DB's from phase one?

On a side note, I think it is great to have a developer here in the fora who is willing to share his insights and views. Thanks for that Andrey!

 

[Robert Campbell]

What I've got so far -- and I don't have a Sony -- is:

1. Advice to shoot at base ISO is/may be wrong -- and if you thought it was 100, well it's something else

2. ETTR is/may be wrong

3. 18% grey is something else

Any other paradigm shifts -- before I give it all up?

 

[Shelby Lewis]

Ok... I'm not going to refute much of what has been said here (from a technical viewpoint)... but here are my observations, so far, from a month's work with the a900 in paying situations (even if that makes no difference ).

I think exposure optimization has more to do with photographic intent than technical means. I use ETTR to keep the amount of shadow noise to a minimum in my photographs to a point. Blindly using it is just being ignorant, if you ask me, but using it smartly to control the dynamic range of a scene is essential at times if you need the most data to work with at the beginning of the post process.

In wedding work, I'm often in situations where the majority of the scene is in deep shadow. I'd rather ETTR to keep chroma noise at bay amap than to lose even more detail by having to deal with excessive noise (dictated by the scene)

I totally agree about the color hit you can take with ETTR... given that the scene isn't all highlights. There are high key scenes that beg for ettr and underexposing deeply and boosting in post (the opposite of ETTR) will (in my book) yield poorer results than an ETTR exposure.

Lastly... and I need to test this more... but 320 is a great ISO on the a900 but i see way to much texture in broad smooth areas (skies, for example) to warrant it's use in many situations. Even if the camera isn't optimized below 320, the camera does take a noise hit at 320... ettr or not.

For me... exposure is just as much a question of intent as it is about optimization... and I make that decision taking a BUNCH of factors into considerations.

Great topic!

 

[hardloaf]

Douglas,

Andrey, I think it should be mentioned that: one, I assume you're speaking of spot metering, and not matrix. Two, sometimes not using ETTR means that highlights may be blown out. I've seen many misunderstand ETTR and not let any highlights blow at all, which results in the underexposure of midtones/subject. Your general advice of making sure that midtones are in the right place sounds good to me.

Yes, for simplicity let's stick to spot metering. Matrix is more complex case, but it's still based around the same gray point as spot metering.

Regarding ETTR I already mentioned this, but let me rephrase - ETTR as universal approach for all kinds of shooting is wrong and unfortunately most of people treat it this way no matter what and how they shoot. It's easy to understand and there is even some technical explanation behind it, but in fact it doesn't tell the whole story.
ETTR as approach when we are trying to open shadows without clipping highlights is a valid technique when needed and as long as we understand what we gain and what we loose there is absolutely nothing wrong with that.

So now about the abbreviation - it's obviously ambiguous and to avoid problems I'd prefer not to use it at all. After all idea to expose for shadows is as old as the world itself, idea that sometimes highlights cannot be clipped is also totally banal, so why do we really need that abbreviation? Just understand what you are doing and expose consciously

 

[Shelby Lewis]

Just understand what you are doing and expose consciously

Preach On, Brother!

That was a much shorter way to say exactly what I said. I should have let you say it first

 

[hardloaf]

Douglas,

Andrey, if you have a scene with very low dynamic range that has no risk of clipping shadows and highlights, where do you place middle grey? -1, 0, or +1 EV? Which is the ideal setting for best color, noise, resolution, etc? I'm debating whether there is any reason for me to be using Zone -1. I've been metering middle grey at +0, or essentially +.5 with Zone turned off. Thanks.

This is again matter of taste - if you used to Zone -1 keep using it, this is just a half stop away from default and it's not enough to cause any major problems with colors.

 

[jlm]

sounds like a review of the zone system would serve many.

with a spot meter, i typically set caucasion flesh tones at Zone VI, one stop over Zone V, V being what my spot meter would record on a neutral gray (18%) card, and what my incident light meter would show.

let the shadows and highlights fall where they may depending on the range of the scene.

one could test their ISO setting by shooting a gray card, metering off the card, and printing it with no exposure change. then compare the gray card to the print.

in practice, it is easier to shoot the gray scale card, still metering off neutral gray.

 

[hardloaf]

I think the 18% came from slide film and color printing paper latitude of 5 f/stops. Slide film had usually a latitude of 2 1/3 over and 2 2/3 under. I also heard that Sekonic grey point is currently at 13% instead of 18%, probably to compensate for the increase in modern film latitude. I think Andrey is making a very valid point about meters. I tested my A900 meter against a Sekonic and the readings were practically similar (if anything, the A900 overexposed very slightly over the Sekonic by 1/10 to 1/20 f/stop). This means that while the meter reading is basically correct, the real iso sensitivity of the camera may not be exactly as advertized. DXOmark suggests about 1/2 stop below advertized numbers.

This discussion is getting more and more interesting

Makes sense. I think it's more like 1/3 stop below though, so shooting 320 means underexposed 250. All vendors got themselves into some funny game - D3X does the same (1/4 of a stop lower), 5D2 (1/2 stop lower). It may look like some marketing ploy, but actually guys are just messing around since there is no standard for ISO in Raw.

 

[hardloaf]

Georg,

Andrey, I realise you are the manufacturer, however, I would think you did make some form of empirical studies comparing your product with the output of other converters, and it was that I was curious about.

In what fields did you find RPP provides an advantage over C1 for example.

It resolves more details, produces better colors and noise and it's much more predictable because it doesn't do anything if not asked to and you have full control over your image. It's different though and past experience with other converters doesn't help at all - no real time preview and workflow should be different to be efficient.

Having said that, I am now shooting with the alpha and of course I am interested in best possible file quality. I am also shooting with a phase one, did you have a chance to try your RPP with DB's from phase one?

On a side note, I think it is great to have a developer here in the fora who is willing to share his insights and views. Thanks for that Andrey!

P20, P20+, P21, P21+, P25, P25+, P30, P30+, P45, P45+ are fully supported and I have very positive feedback from those who work with them. You need a fast computer though, Mac Pro is highly recommended for everything above 16M.

 

[Georg Baumann]

<grins> I am installing a 8 core 2.66 Ghz Nehalem with 32 Gig Ram as I type here, that should do.

Thanks Andrey, no real time preview means exactly what? I change parameters xyz and do not get to see the effect in the preview?

<sorry, I should really DL your RPP and give it a go before asking questions. :ROTFL:>