PhaseOne P25 ISO test. 200/400/800

[brumbaer]

I will not add to the discussion about 35mm and MDFB it's a waste of time as all religious wars, but I have 5 cents about noise reduction.

Many things you do reduce noise, some are inflating it. If you want the least (absolute) noise reduction applied, you will have to open your camera and remove any filters from the PCB (happy soldering), avoid downscaling your image, avoid reducing the bit depth or reducing the levels, prevent application of the blackfile and so on. This is in some cases not possible. So whatever you do,  some noise reduction will have been applied. Your NEF is not "raw".

What you will get in a NEF is less raw than what you get from an IA/BR file combination, because a lot of processing has been applied to the NEF in camera.

When you talk about noise reduction and sharpening you probably talk about dedicated algorithms which reduce noise and increase sharpness. The kind which will create artifacts when excessively applied.

eMotionDng does neither apply any noise reduction nor sharpening to the image in the sense of algorithms targeted to do so. You can reduce noise for the whitefile, which doesn't reduce the noise on the image. The DNG created by eMotionDng is at least as unprocessed as the NEF.

Of course I can't speak for any other software. And I do not know what processing is automatically applied by ACR, Lightroom or whatever.

regards
SH

[rethmeier]

Ray,
Graham uses the eMotion 54LV not the eMotion 75LV.
Cheers,
Willem.

That's why I made the comment!

[thsinar]

hi Willem,

FYI: Graham has used "my" eMotion 75 "ISO 800" image sample to make his point.

Kind regards,
Thierry

Ray,
Graham uses the eMotion 54LV not the eMotion 75LV.
Cheers,
Willem.
[a href=\"index.php?act=findpost&pid=197395\"][{POST_SNAPBACK}][/a]

[thsinar]

just for those who are not familiar with the eMotion backs and its files (and I believe "Panopeeper" is not):

a ".IA" = Image Archive and a ".BR" = Black Reference are the 2 components of any eMotion RAW file (as raw as it can get, as mentioned by Stefan).

One has the unique possibility to get access to these RAWs, when using a eMotion back.

For the rest, I guess it makes it clear, after Stefan's explanation, that the Brumbaer DNG converter does not apply any NR, as implied. And the explanation of noise reduction (denoise) of the white shading is the one I have given on another forum: it does de-noise the white shading, but leaves the image data as it is/was.

Best regards,
Thierry

What you will get in a NEF is less raw than what you get from an IA/BR file combination, because a lot of processing has been applied to the NEF in camera.

eMotionDng does neither apply any noise reduction nor sharpening to the image in the sense of algorithms targeted to do so. You can reduce noise for the whitefile, which doesn't reduce the noise on the image.

SH
[a href=\"index.php?act=findpost&pid=197393\"][{POST_SNAPBACK}][/a]

[EricWHiss]

Thierry,
I still have the ISO 800 sample file you made available in another post and after I did my tests with my own back, the P20, I did compare the results to that e75 file.  There is no question that the e75 has less noise than my p20.  No argument at all there at all.  Getting back to the original point of the thread,  I am quite amazed that my older generation back actually can do quite well, much better than I would have guessed.  All of this is good news for MFDB users.  

Eric

[thsinar]

Yes, Eric, I absolutely agree with you.

I don't know who did misguide (if somebody did) or where from this belief comes, but my primary intention with this test done a few weeks ago was to make exactly this point: that MFDBs are not that bad at all at higher ISOs: Whatever we call ISO, however we achieve it, the important is ultimately the end resulting image.

Best regards,
Thierry

Thierry,
I still have the ISO 800 sample file you made available in another post and after I did my tests with my own back, the P20, I did compare the results to that e75 file.  There is no question that the e75 has less noise than my p20.  No argument at all there at all.  Getting back to the original point of the thread,  I am quite amazed that my older generation back actually can do quite well, much better than I would have guessed.  All of this is good news for MFDB users. 

Eric
[a href=\"index.php?act=findpost&pid=197398\"][{POST_SNAPBACK}][/a]

[rethmeier]

hi Willem,

FYI: Graham has used "my" eMotion 75 "ISO 800" image sample to make his point.

Kind regards,
Thierry
[a href=\"index.php?act=findpost&pid=197396\"][{POST_SNAPBACK}][/a]

Hi Thierry,
that was the reason for my post reply!
Cheers,
Willem.

[rethmeier]

Also,
I think it's time to move on from this!

At the end of the day,all it counts how it looks printed in a high-end coffee table book.

And from experience a 5D will give great results for a single page!

Why do I use a Hy6/e-75LV as well?

Because I love  and can afford it!

Cheers,
Willem.

[Ray]

Yes, Eric, I absolutely agree with you.

I don't know who did misguide (if somebody did) or where from this belief comes, but my primary intention with this test done a few weeks ago was to make exactly this point: that MFDBs are not that bad at all at higher ISOs: Whatever we call ISO, however we achieve it, the important is ultimately the end resulting image.

Best regards,
Thierry
[a href=\"index.php?act=findpost&pid=197402\"][{POST_SNAPBACK}][/a]

Thierry,
There seems to be some confusion here (in my mind at least). According to Sinar's website, the 22mp eMotion 54LV has an ISO range from 50-200 and the 33mp 75LV, 50-400.

Do I take it that the ISO 800 sample image is actually ISO 200 underexposed by 2 stops?

At the risk of laboring a point, if the 22mp 54LV were compared with the 1Ds3 for the purposes of noise comparison, it would only be sensible to use one stop lower ISO with the 1Ds3 (after having adjusted FL and aperture for same FoV and DoF, of course).

In other words:
(1) the 54LV underexposed 2 stops at ISO 200 compared with the 1Ds3 correctly exposed at ISO 400.

(2) the 54LV underexposed 3 stops at ISO 200 compared with the 1Ds3 correctly exposed at ISO 800.

(3) the 54LV underexposed 4 stops at ISO 200 compared with the 1Ds3 correctly exposed at ISO 1600.

By correctly exposed, I mean an exposure which pushes the histogram as far as possible to the right without clipping highlights.

[samuel_js]

Thierry,
There seems to be some confusion here (in my mind at least). According to Sinar's website, the 22mp eMotion 54LV has an ISO range from 50-200 and the 33mp 75LV, 50-400.

Do I take it that the ISO 800 sample image is actually ISO 200 underexposed by 2 stops?

At the risk of laboring a point, if the 22mp 54LV were compared with the 1Ds3 for the purposes of noise comparison, it would only be sensible to use one stop lower ISO with the 1Ds3 (after having adjusted FL and aperture for same FoV and DoF, of course).

In other words:
(1) the 54LV underexposed 2 stops at ISO 200 compared with the 1Ds3 correctly exposed at ISO 400.

(2) the 54LV underexposed 3 stops at ISO 200 compared with the 1Ds3 correctly exposed at ISO 800.

(3) the 54LV underexposed 4 stops at ISO 200 compared with the 1Ds3 correctly exposed at ISO 1600.

By correctly exposed, I mean an exposure which pushes the histogram as far as possible to the right without clipping highlights.
[{POST_SNAPBACK}][/a]

No, the are referring to the new version (iso 100-800)
[a href=\"http://www.sinarcameras.com/file_uploads/bibliothek/k_92_Brochures/436_0_prosphy6_210x280_gb_150dpi.pdf]http://www.sinarcameras.com/file_uploads/b...0_gb_150dpi.pdf[/url]

Pleae Ray, get over it. It's friday!  

[amsp]

No don't mind at all and thanks! Would not have tried this were it not for your post.   

I did just now try the same image in C1 3.8 DB  The B&W conversion using the B&W film looks and also using the panachromatic color profile looks good, but I think now seeing both I can say the noise handing of C1 4.1 is better.   There is are a few vertical bands in the image in 3.78 DB that I didn't notice in 4.1   

amsp, which do you prefer  3.78 or 4.1?
[a href=\"index.php?act=findpost&pid=197324\"][{POST_SNAPBACK}][/a]

So far I prefer 3.8, but I haven't done extensive testing.

[eronald]

In my experience, the issue with the Phase backs is that they stripe at higher ISO. This is possibly a reflection of lacking hi-iso calibration. Noise is really not the problem. Under-exposing with a lower ISO in my experience (or imagination) seems to minimize this striping.
Quite possibly the problem will at some point be brought under control by software, as competitors advance. When I raised a ruckus about bad 400 ISO with my first P45+ back the Phase response was that my images were underexposed, and the dealer explained that I shouldn't be shooting at high speed. It's pretty obvious now that the attitude of the competition has changed, and ISO 800 is now given as shootable by Sinar. Frankly, if Sinar or Hassy can do high ISO well, I will be switching to one of these brands in due course.

Edmund

[Guy Mancuso]

Guy, 
I hope you will post some examples from the Horseman either here or on your site!

Thanks
Paul C
[a href=\"index.php?act=findpost&pid=197301\"][{POST_SNAPBACK}][/a]

I get the Horseman today so will be working on some shots to post. I have a loaner P25 until my P25 plus comes in early next week.

I just wanted to add something here. First off we are not comparing it to DSLR's and the original intent from the OP is hey guys look we can actually use the higher ISO settings on these backs. Something maybe we could not have done before but after using the Phase backs you get really clean files at the 400 , 800 settings as well as the Sinars. Now I have shot digital from day one 15 years or so ago and had every system made , Nikon, Kodak , Canon , Leica and they have there place and yes you can pull some higher ISO' than the DB like the D3 for instance it is very clean. So bottom line you need ISO 3200 or higher buy a Nikon but whoever thinks for a second a DSLR is better file wise in detail , micro contrast and such is sniffing something he should not. And frankly you sound pretty stupid even suggesting they are. I have had them all and tested them all and MF is a eye opener , simply put bigger is better. Frankly the only thing close was the DMR and M8 but there issue is there a few mpx short in the pants. Honestly the only true way to find out is go drop your money down and buy one.

[woof75]

This is fascinating, if I keep reading this thread I will know for sure which is better, 35mm or medium format. What will the answer be, I can't wait.

Come on if you don't have actual information or an actual question just knock it off, endless silly arguments by people coming up witH hypothetical comparisons as to what might be theoretically better is no use to anyone.
Lets end it here, there is a price and that is that people don't feel safe to post these type of examples of high iso oR whatever for fear of the thread being dragged into the gutter.

THIS IS A PLACE TO SHARE KNOWLEDGE NOT PROVE YOUR AGENDA!!

[Ray]

No, the are referring to the new version (iso 100-800)
http://www.sinarcameras.com/file_uploads/b...0_gb_150dpi.pdf

Pleae Ray, get over it. It's friday! 
[a href=\"index.php?act=findpost&pid=197411\"][{POST_SNAPBACK}][/a]

Okay! Understood. There's a new model with an ISO 800 capability. Since that's 33mp the situation changes somewhat. Downsizing 33mp to the 21mp of the 1Ds3 changes slightly the DoF relationship. By how much it's not certain.

However, it seems clear to me that ISO for the same shutter speed, DoF and FoV, whilst not as much as one stop difference, will still not be the same. It will be higher for the e75, perhaps by 2/3rds of an F stop. We need to see the comparisons to find out just how much.

But let's not quibble. Let's see a competent comparison so we can see just how significant the improvement is.

Those who argue this is a religious war between MFDB and 35mm are totally wrong. I'm not religious in the slightest degree. I'll always try to bend with the facts.

Unfortunately, I'm not getting much in the way of facts from you people. Most of your images are of heavily made up models trying to create an illusion of desirability. The desirability of the subject matter is clearly connected to the desirability of the equipment

The fact that you cannot produce a competent comparison between a 1Ds3 and a P25 or eMotion 54LV indicates that you are truly in the business of illusion and don't care about the facts.

The accusation by highly regarded peple like James Russell that I'm unreasonable and obsessive because I expect busy professionals to take the trouble to do proper comparisons, is totally misplaced. I don't expect that. If you are too busy, I understand. There's no problem there at all for me.

All I ask is, if you are going to take the trouble to make comparisons, then do it properly.

As my mother always said, if a job is worth doing, it's worth doing well. Don't you guys follow that principle?

I'm not sure I trust you guys. The following shot is not of Owls but Frogmouths.

[attachment=6729:attachment]

[Graham Mitchell]

The fact that you cannot produce a competent comparison between a 1Ds3 and a P25 or eMotion 54LV indicates that you are truly in the business of illusion and don't care about the facts.

Ok, you are starting to get on my nerves now. I don't happen to have a spare 1Ds3 lying around. Send me one and I'll gladly do the test.

[Ray]

Ok, you are starting to get on my nerves now. I don't happen to have a spare 1Ds3 lying around. Send me one and I'll gladly do the test.
[a href=\"index.php?act=findpost&pid=197444\"][{POST_SNAPBACK}][/a]

I don't either. You think I'm concerned with brand wars here? I'm concerned only with photographic principles. I want to cut through the bull***t.

You don't have a 1Ds3 and I don't have a DB, of any sort. You are more likely to be able to afford a 1Ds3 because you've paid $30,000 for a DB and are a professional photographer with a revenue stream.

I'm an amateur who is very concerned about image quality, but I'm reluctant to waste money hiring equipment at a very high cost in order to confirm (or debunk) what my reason tells me is true.

[woof75]

I don't either. You think I'm concerned with brand wars here? I'm concerned only with photographic principles. I want to cut through the bull***t.

You don't have a 1Ds3 and I don't have a DB, of any sort. You are more likely to be able to afford a 1Ds3 because you've paid $30,000 for a DB and are a professional photographer with a revenue stream.

I'm an amateur who is very concerned about image quality, but I'm reluctant to waste money hiring equipment at a very high cost in order to confirm (or debunk) what my reason tells me is true.
[a href=\"index.php?act=findpost&pid=197450\"][{POST_SNAPBACK}][/a]

Please don't feed the trolls!

[Ray]

Please don't feed the trolls!
[a href=\"index.php?act=findpost&pid=197460\"][{POST_SNAPBACK}][/a]

Please try to make an intelligent comment.

[BJL]

As Ray well knows, there are some advantages of larger formats over smaller ones that, for some photographers in some photographic situations, make a larger format the only acceptable choice, regardless of any advantages of smaller format gear is other respects. In that situation, the only noise level assessments of practical relevance are of the acceptable options; in this case, the various medium format options.

I would guess that medium format is often chosen simply for advantages like superior sharpness and resolution with currently available sensor and lens combinations. And once this is the case, comparisons to smaller format options are irrelevant, while assessments of the acceptable (MF) options are of far greater interest. Parts of this thread are as if someone were to post over and over again advocating the size and weight advantages of 4/3 format kits like an Olympus E-420 with 14-42/4-5.6 and 40-150/4-5.6 over 35mm format kits like the Canon 5D, 100-400/4-5.6, etc. that Ray uses, despite the clear unacceptability to Ray of that 4/3 format option for other reasons.