A900 DRO ???

Specifically for the discussion of the A-mount DSLR range
Forum rules
No more than three images or three external links allowed in any post or reply. Please trim quotations and do not include images in quotes unless essential.
User avatar
ianmiddy
Heirophant
Posts: 106
Joined: Sun May 20, 2007 8:26 am
Location: Derby, England & SW Scotland
Contact:

A900 DRO ???

Unread post by ianmiddy »

David [and any other 900 owners] - have you tried Advanced DRO on the 900 yet ?

Curious how your experience compares with the 700 [ref the Photoworld articles on it, and also its use with flash exposure - eg 2008#1] - has the new sensor made any difference to how its been implemented, or the type of result at the different [1-5] settings ?

[obviously, most interested in JPG, as I assume the RAW/PP implementation is effectively unchanged ?]

Cheers

IDM
David Kilpatrick
Site Admin
Posts: 5985
Joined: Sat May 19, 2007 1:14 pm
Location: Kelso, Scotland
Contact:

Re: A900 DRO ???

Unread post by David Kilpatrick »

No full test, but I shot quite a lot with DRO+3 and it seemed exactly the same as the A700. They do claim is improved, but v4 firmware on the A700 is also supposed to improve DRO+. I will check it properly when I get my A700 back :-)

David
User avatar
ianmiddy
Heirophant
Posts: 106
Joined: Sun May 20, 2007 8:26 am
Location: Derby, England & SW Scotland
Contact:

Re: A900 DRO ???

Unread post by ianmiddy »

Thanks, David - did I also read somewhere that the 900 does away with the [Minolta] Hi200/Lo80, whilst the 700 kept them alongside the DRO implementation ?

Cheers

IDM
David Kilpatrick
Site Admin
Posts: 5985
Joined: Sat May 19, 2007 1:14 pm
Location: Kelso, Scotland
Contact:

Re: A900 DRO ???

Unread post by David Kilpatrick »

Correct. There is no Hi/Lo option on the 900, they assume I guess that DRO will take care of this aspect. DRO Standard is similar to Hi and Lo combined.

NB: the Nikon D90 warns, again, in its manual that using Active D-Lighting will result in underexposed raw files and may increase noise - a warning which Sony fails to make about DRO+.

David
User avatar
pakodominguez
Minister with Portfolio
Posts: 2306
Joined: Tue May 22, 2007 5:38 pm
Location: NYC
Contact:

Re: A900 DRO ???

Unread post by pakodominguez »

David Kilpatrick wrote:Correct. There is no Hi/Lo option on the 900, they assume I guess that DRO will take care of this aspect. DRO Standard is similar to Hi and Lo combined.
David
Yes, they keept the the Zone Matching (Hi/Lo) function, you can set it up in the Custom Creative Style options. It's call just "Zone" now and, in words of my favorite Sony rep, "it's a Minolta thing" (...) see page 80 to 82 in the Manual: ftp://ftp.vaio-link.com/pub/manuals/con ... 682121.PDF

Regards
Pako
------------
http://www.pakodominguez.photo/blog" onclick="window.open(this.href);return false;
User avatar
ianmiddy
Heirophant
Posts: 106
Joined: Sun May 20, 2007 8:26 am
Location: Derby, England & SW Scotland
Contact:

Re: A900 DRO ???

Unread post by ianmiddy »

pakodominguez wrote:Yes, they keept the the Zone Matching (Hi/Lo) function, you can set it up in the Custom Creative Style options. It's call just "Zone" now
Ah, thanks, Pako - I assume this is the same as the A700 [looking at my brochure for it, as I still can't afford a 'real' one!] as that seems to suggest that 'brightness' & 'zone' were added to the KM5D/A100s Con/Sat/Sharp settings [whatever happened to the 7D's 'Hue' ?]...

Cheers

IDM
David Kilpatrick
Site Admin
Posts: 5985
Joined: Sat May 19, 2007 1:14 pm
Location: Kelso, Scotland
Contact:

Re: A900 DRO ???

Unread post by David Kilpatrick »

It is not the same as Zone Matching or Hi Lo - it is a gamma curve adjustment instead. It controls the midtones along with the brightness adjustment on the A700/A900 and does not affect the exposure or ISO values, or the retention of highlights and shadows. They probably changed the way this works because DRO+ can be used with Brightness and Zone adjustments (which replace Contrast adjustments when active - you can not also adjust contrast).

NB - using Zone on the A700/900 can be very misleading if you use the rear screen to check your exposures. The histogram always shows the JPEG, and Zone only affects the JPEG. You may see apparently normal skin tones etc if you set a bright zone setting, when actually the shot is much darker. I made this mistake for a short period. Only use it if the JPEG out of the camera is what you want for your final shot.

David
Javelin
Emperor of a Minor Galaxy
Posts: 1856
Joined: Wed Jul 16, 2008 4:51 pm

Re: A900 DRO ???

Unread post by Javelin »

I thought the zone matching affected the raw exposure as well as the jpg. and only DRO didn't affect the raw?
User avatar
ianmiddy
Heirophant
Posts: 106
Joined: Sun May 20, 2007 8:26 am
Location: Derby, England & SW Scotland
Contact:

Re: A900 DRO ???

Unread post by ianmiddy »

David Kilpatrick wrote:DRO+ can be used with Brightness and Zone adjustments (which replace Contrast adjustments when active - you can not also adjust contrast)
Ahhh - this would be so much easier to visualise/understand if I had a camera to try combinations out on [and then maybe remember, lol] - oh well, hopefully the cashback may continue after tomorrow and I might get one soon...

Thanks, all, for the pointers.

IDM
David Kilpatrick
Site Admin
Posts: 5985
Joined: Sat May 19, 2007 1:14 pm
Location: Kelso, Scotland
Contact:

Re: A900 DRO ???

Unread post by David Kilpatrick »

Hi200 and Lo 80 (D5D-A100 etc) both use an EI 125 sensor gain, but expose (metering) as if the sensitivity was 80 and 200 respectively - raw files are affected, by just by getting more or less exposure, and not by the tone curve which creates the JPEGs. Like DRO, the effect is present only in the in-camera JPEG or the IDC developed file. However, because Lo80 is like a raw file with +0.6 exposure over-ride, you get better shadows and just have to recover highlights or hope that none are burned out; and Hi200 is like a -0.6 over-ride, meaning you are less likely to have burned out highlights.

The Brightness and Zone controls of the A700 and A900 do not alter the ISO setting, and can be used at any ISO (though I think that like DRO, the effect is turned off once you set above ISO 800). So the raw files from the A700 and A900 will be just the same as regular raw files in density. The big difference in practice is that the D700 and A900 rely on the improved dynamic range and headroom present in the raw file, instead of deliberately underexposing as Hi200 does. Since DRO+ will underexpose as needed to get the right density of file to work with, it would be double-damage if any other function was also adding underexposure.

The Konica Minolta quote Don provides is marketing hype - the modes do NOT extend the CCD dynamic range, that remains very much fixed! What they do is adjust the relative exposure to make better use of the CCD dynamic range, and then apply special tone curves to produce a good-looking JPEG from this adjustment. If they could extend the CCD dynamic range.... well, they would do it permanently and not have a special mode. Yes, I know that Fuji's system has a special mode to extend dynamic range (it's different because this adjusts a blend of two pixel types and their relative input to the final pixel values) and that Kodak's ERI-JPEG also allowed a dynamic range change, but that was effectively like switching bit depth. Also, Nikon 14-bit raw tends to produce a better dynamic range than 12-bit raw.

That is not what KM did (or Sony inherited). Lo80 and Hi200 are simple, clever uses of exposure adjustment and tone curve for in-camera JPEGs.

David
David Kilpatrick
Site Admin
Posts: 5985
Joined: Sat May 19, 2007 1:14 pm
Location: Kelso, Scotland
Contact:

Re: KM5D Zone Matching on Raw

Unread post by David Kilpatrick »

Sonolta wrote:
Again, not so fast...open these two untouched .MRW files in ACR and you will clearly see a MONSTER difference between Hi 200 and ISO 200 on Raw! The *Effect* is clearly visible on the ACR developed Raw file, and most certainly you can see a MAJOR benefit of shooting HI 200 Raw! Tweak the files for yourself if you must!

Sure - Hi200 is ISO 125 underexposed by 0.6 stops, and ISO 200 is ISO 200 exposed normally. Of course there's a difference in raw density. But Hi200 is not actually any different to ISO 125, exposed with -2.3rds compensation - there is no tone curve applied to the file, as there can not be. KM's guys explained how all that worked to me back in 2004/5. It is a simple process, but the raw file is just a normal raw file given a biased exposure.

The point about Hi200 is that it is not ISO 200 - it's an underexposed ISO 125 which is subsequently processed in camera to brighten the midtones, but leave the highlights that little bit darker (as the tone curves you showed indicate).

(I was told it is 0.6 at 125 - it may be 1 stop at 100, but my info is that both Lo80 and Hi200 use the same 'optimum' sensor gain which happens to be 125 in the 5D, and was 160 in the 7D, hence Hi250).

David
David Kilpatrick
Site Admin
Posts: 5985
Joined: Sat May 19, 2007 1:14 pm
Location: Kelso, Scotland
Contact:

Re: A900 DRO ???

Unread post by David Kilpatrick »

Don, it doesn't alter the way this is achieved - by either underexposing a normal raw file and using a changed conversion curve, or overexposing a raw file and doing the same.

The difference with the A700 and A900 is that curves are applied to normal rather than biased exposures, giving a choice of ISO settings which is not present in the 7D to A100 method.

I know you will be able to understand this. I'm not denying the value of the settings, but it is important to understand what they are doing. This was explained to me (and all others) by the Japanese tech designers at the original 7D launch photokina and subsequent conferences. All the function curves you show are from JPEG conversions, because that's how the site involved measures the values.

David
David Kilpatrick
Site Admin
Posts: 5985
Joined: Sat May 19, 2007 1:14 pm
Location: Kelso, Scotland
Contact:

Re: A900 DRO ???

Unread post by David Kilpatrick »

You still don't understand? You must have been popular in class.

The headroom is simply utilised by controlled underexposure or overexposure of the raw file, and a special JPEG curve.

If you shoot with zone matching and do not use the JPEG, all you actually get is a light or dark raw file (by -0.6 or +0.6 EV) at a specific ISO rating - 160 in the D7D, 125 in the D5D and A100.

If you then process the raw file without any adjustment in ACR (etc) it is simply like applying a similar plus or minus exposure correction. IDC, or Dimage Master, can apply curves similar to the in-camer conversion but ACR can not unless you create a curve to match the transfer functions of the zone matching in-camera. You will not get a tonal adjustment from within the raw file itself, just a lightening or darkening. Nothing that you have flagged up is inconsistent with this, and nor need it be, because what I am telling you is how it works.

Replying to this does not need bold type, or capital letters, or GESSH or exclamation marks. That's just rude and silly and childish. You are wasting my time, and everyone else's, by arguing some point that does not even need arguing. Zone matching works - I have never claimed it did not - but if you want to real, full benefit of the enhanced transfer function curves this can only be obtained with in-camera JPEGs - just like DRO, which came after zone matching (and alongside it, though you will note, never together for reasons I have explained).

What I don't get is - why do you do this? Do you have a dysfunction of some kind? What makes you want to argue that black is white so often?

David
David Kilpatrick
Site Admin
Posts: 5985
Joined: Sat May 19, 2007 1:14 pm
Location: Kelso, Scotland
Contact:

Re: A900 DRO ???

Unread post by David Kilpatrick »

Don, repeat your test and set -0.6 exposure for your ISO 200 raw file. It will not be entirely correct, since it will be slightly more noisy when processed than Hi200 (it will, in effect, be Hi320).

There is absolutely no benefit in RAW shooting to using Hi or Lo Zone Matching, relative to using similar levels of minus or plus exposure over-ride. Doing that will give exactly the same recoverable detail provided you can set 1/3rd stop accurate ISO - the sensor gain is carefully optimised to EI 125 (overexposing ISO 100 will not work as well, since saturation is reached, and underexposing 200 won't either - but overexposing 200, and underexposing 100, will produce results which may even be marginally better than Lo80 and Hi200 respectively).

Zone Matching involves applying a simple transfer curve which, for example, may take what should be a value of 245/245/245 (as normally converted from raw) and will render it as 240/240/240 in JPEG (Hi type example).

Please examine your own histograms;

Image

Note the bottom end of the histogram, and the position of the three sensitivity peaks which indicate it is not exposed in daylight conditions. The black level is not clipped, the highlight end is clipped.

Image

This is an identical histogram except for an exposure shift; the black level is now clipped, to exactly the same degree as the highlight end is moved away from clipping (though it is still clipped, and ACR highlight recovery would be needed to pull this in).

All you have shown here is a simply exposure shift - a normal 200 ISO shot is a normally exposed ISO 200; a Hi200 shot is the same exposure of f/9 at 1/15th, but the ISO has been 'secretly' changed to 125 although ACR still reports 'ISO 200'.

To see the true state of the exposure, set the black level to 0 in both cases, then reduce the exposure slider until the base of the shadow detail gradient ends neatly at 0 (as it does with black clipping set to 5 on the standard example).

All that your two raw files is show that one exposure is less than the other - where highlight detail is gained, shadow detail is lost, as normal. Try Hi200 and normal 200 JPEGs from the camera, and you will see a different story. It really, honestly, truly is necessary to use in-camera JPEGs to gain the benefits of Hi200 and Lo80 and it is not me that's spreading misinformation and confusion - it is you.

I'm off to bed anyway. I spent today delivering a seminar lecture on raw processing to a group of professional photographers who displayed excellent comprehension of every aspect, and coming back to have to deal with your refusal to even engage the brain and understand the process is frustrating because I know that once you understand what's happening, it will be useful to you.

David
david antony
Viceroy
Posts: 1213
Joined: Wed Jun 11, 2008 8:32 pm
Location: tsawwassen, bc. canada.

Re: A900 DRO ???

Unread post by david antony »

Sonolta put your toupée back on before it runs away <|;o)
Post Reply

Who is online

Users browsing this forum: No registered users and 51 guests