Magic Lantern Improves 5D Mark III Dynamic Range to 14 Stops

[SubPrime]

qianp2k wrote:

schmegg wrote:

I understand where you're coming from based on who you were replying to initially. ;-)

Or in one year's ago in his tons of D800 vs 5D3 posts in this forum, so I am not surprised he has special passion in his Nikon gear ;-)

I have passion for both ;-)

It's like having two kids who I love equally. I won't hear a bad word said about either one.

But saying it is significant because you use both holds no more or less weight than another dual system user who says it isn't significant. Plain fact.

I'd rather to believe well established photographers such as Hans and Galland who also own both systems or used to use both, and both said there is no significant difference. Nobody would suggest them a fanboy in one system or another.

I suspect you would rather to believe the word of someone who agrees with you and makes you feel a little less insecure, in spite of the fact you have no reason to be. The 5D3 is more practical is many cases precisely because it has lower resolution - that doesn't make it less a camera.

I have already pointed out that Hans is not impractical. I am not familiar with Galland's comments.

 

[qianp2k]

Phil M Winder wrote:

qianp2k wrote:

Phil M Winder wrote:

qianp2k wrote:

Phil M Winder wrote:

qianp2k wrote:

Phil M Winder wrote:

qianp2k wrote:

Phil M Winder wrote:

qianp2k wrote:

Marquee wrote:

ETTR means "Expose to the Right", so looking at the histogram, the exposure is as far right as possible without clipping highlights - in reality you allow the highlights to be clipped a tad because the exposure information you get is derived from a Jpeg representation of the RAW you took (assuming you are exposing in a RAW mode), so some clipping can be recovered later. A RAW file that can then take significant pulling of shadow detail could then repsesent a large dynamic range albeit in a form within the viewing limitations of the protraying medium, being monitor or print. Or you could blend two exposures of course which in many cases will be equally good or better (assuming a static subject)

Here are two BIF samples I used ETTR technique. It's under bright early afternoon sunlight at Cancun beach. I was shooting pelican in flight against bright sky and ocean wave. I used Partial metering which is a larger version of Spot metering so already gain +1/3 EV from normal Evaluative metering (that similar to Nikon's Matrix metering).

I gave +2/3 to +1 EC (exposure compensation) on top of Partial metering. So the main subject, the pelicans, were exposed correctly so I don't have to pull shadow aggressively. In OOC RAW files, the highlight, sky or ocean waves were a bit of overblown but I still can recover in LR4 (Recovery bar, software GND filter and Brush tool if necessary). Someone please shows us D800/D600 ETTL (expose to left) BIF photos that exposed on sky then pull birds from deep dark shadows, and let's compare feather details at large size :-D

--
http://qianp2k.zenfolio.com/

If you are exposing for the main subject, the birds, then why would you want to compare with D800/D600 BIF photos that expose for the sky?

That's why I am asking and waiting to see. I saw someone did before that shows obvious side effect - detail lost and noisy/grainy (sure you can apply NR but that further smear details). If someone doubt this ETTR technique then the burden falls to that person not me to prove why ETTL would do better.

Also, where are you deciding that you should increase to +2/3 or +1 EV?

a) experiences; b) checked onsite on back LCD - histogram and 10X playback (5DIII LCD works very well even under bright light) and adjusted accordingly, that how I did in airshows as well.

http://www.dpreview.com/forums/thread/3238329

--
http://qianp2k.zenfolio.com/

What are you trying to compare if you are both exposing for different things?

That's why I am also asking BIF thru deep shadow lifting method.

Why not ask for same method?

The same ETTR method would not make much difference as clearly shown in Rick's side by side test between 5D3 and D800 with Zeiss 35/f2.0 prime, and it's my suggested method. I only want to see as someone keep saying ETTL would generate better result or no penalties

But if you are exposing for the subject (the deep dark bird in this case) you aren't really quite ETTR, are you?

Then it's metering related, right? I used Partial metering that is a bigger version of Spot metering. here is a shot against black bear in shadow that itself against bright background - Partial metering and no EC and 1D3 metering spot on. EC is further compensation beyond metering. So doesn't matter black or white birds.

ETTR is not necessarily related to metering.

Yes as I said, so not sure what black birds related to ETTL but metering.

The image, like the birds, do not show a situation that could benefit from ETTR.

It did as I intentionally overexposed around one stop so birds got sufficient exposure that more than normal exposure. OOC raws show highlight a bit overblown but I successful recovered them that result cleaner, more details and sharper birds, the main subject, although small subject.

I'm asking because it does not seem like the situation called for "ETTR" or "ETTL". Arbitrarily adding or decreasing exposure comp is not ETTR or ETTL.

ETTR or ETTL is determined by histogram, to right call ETTR to left call ETTL. I changed not arbitrarily BTW.

How did you determine the point of where your highlights or shadows are clipped?

In histgrams onsite, and further details in software.

Do you mean after you have taken the shot?

a) first based on experience; b) after a few test shots.

Ok. after the shot.

After test shots. I know you can boast EVF advantages but I am one of them cannot to live with current EVF that is bad in tracking moving subjects anyway.

Experiences? How were you able to track a deep dark BIF, in the bright afternoon sun and ocean waves, while checking 10x playback on your LCD?

Yes 5DIII LCD works very well as DPR review said, try yourself you will know, lots better than your E-M5 BTW. From LCD I learned my exposure from histogram and still can check details thru 5x/10x playback even under bright sunlight that I know you don't experience at similar level on your E-M5 LCD ;-)

Whether you think the E-M5 LCD is better or not is irrelevant. I didn't expect a logical answer though, as it would be quite difficult to compose a shot of a BIF while checking your LCD.

But obviously I did. But birds or airplanes or lions in my forthcoming Africa safari are not runing in front my eyes every second ;-) Doesn't matter you shoot with whatever camera including E-M5 you still check results on back LCD not in your EVF at least.

Same question. Do you mean after you have taken the shot?

Ditto above. Please don't tell me EVF is better ;-)

Ditto above.

BTW, are you interesting to buy a 5DIII as you seem have special interest on this camera? All I can tell you is a fantastic camera with some best lenses in DSLR world :-D

I have interest in many types of cameras.

Your posting history seem suggesting you particularily interest in Canon cameras in this forum. 5DIII certainly is lots better BIF or airshow camera than mirrorless in general :-D

Certainly is better....if BIF or airshows were a concern.

Better in every aspect, which one is not better for example?

Not better in many aspects. And there certainly isn't just one.

What exact aspects if you can give? Your E-M5 is not even close to 5DIII in IQ and performance.

70-200L/4.0 IS on 5D3 vs 70-200L/2.8 IS II on 5D3 vs Panny 35-100/2.8 OIS on GH2 - 18 vs 21 vs 7 p-mpix. Gee, $1100 Panasonic Lumix G X Vario 35-100mm/f2.8/power OIS zoom with FF eq 70-200mm AOV and F5.6 DOF and total light gathering is pity at 7 mpix. Even DXOmark ever tested on EM-5 or GH-3, it will only gain max 1 mpix to total 8 mpix, far far below two Canon 70-200L IS zoom on 5D3. 5D3 is far better not only in resolution/sharpness but in color tonality and in noise/grain cross entire ISO range that especially oblivious in high ISOs. Can you show us a 100% of E-M5 photo for example? :-D

--
http://qianp2k.zenfolio.com/

 

[SubPrime]

qianp2k wrote:

SubPrime wrote:

qianp2k wrote:

http://www.dpreview.com/forums/thread/3514784

You got kidding me. How many dual system owners we have heard said virtually no difference when print to 30x20" or even 40x30".

You came up with one name. I am a dual system owner who says you can see noticeable difference when print to 30x20"

You must kidding me that if you print to 20x30" from above Rick's samples, you will see noticeable difference?

I wasn't referring to that example.

There are other comparison which reveal that the D800 renders detail the 5D3 cannot even detect and no amount of up-sampling or post processing tricks will change that.

As I pointed out to you in another comparison between a D800E and an IQ180, there will be cases when the higher res superior system does not display any noticeable advantage.

I doubt you will even see noticeable different at 30x40".

You can doubt it, or you can compare actual prints as I have done.

I upsampling 5D3 file to match to D800 size and would have pixel peeping to see very small difference. I have a few 20x30" prints from 5D1 and 5D2 and honestly I even don't see much difference either. The difference between D800 and 5D3 is far less than between 5D2 and 5D1 according to DXOMark system test (lens+sensor).

I read bunch of photogeraphers such as Hans Kruse who truly don't have agenda and widely respectied photographer said they don't see much difference. I'd rather believe him than you.

Again, I have pointed out that Kruse was happy to exaggerate the difference between the D8002 and HD-40, which shows even less disparity than the D800 and 5D3 comparisons. Even as the the reviwer, Alex Koloskov, said the difference would not be noticeable on even a large print, Kruse insisted that the differences were significant.

That makes Kruse biased IMO.

My processed samples are directly out of CS6 by upsampling 5D3 file to D800E size and all others are the same default/Zero setting that everyone can duplicate. Or you processed D800E RAW and I process 5DIII RAw and let's compare?]

To be honest you're strongly Nikon biased from many have seen in your posting history.

I would disagree. If you have looked at my posting history, you will notice that on the Nikon forums, I have been a fierce critic of Nikon, and even been dismissed as a Canon fanboy.

And are you seriously going to deny that you are not only strongly Canon biased, but entirely?

Sure I only shoot Canon but I have seen enough D800/D600 photos and I don't see they are much better in IQ overall.

But don't you realize how weak that sounds when you are comparing your limited experience with D800/D600 photos to someone who shoots both systems all the time and in all manner of

I have read others who widely regarded as honesy and not brand binded as Galland (may not spell his name correclty). He is 5D2->D700->D800->5D3 shooter and he has true professional bussiness. He also said there is no much difference and he switched back to 5D3 because it has better AF.

I would like to read Galland's comments and find out what kind of shooting he does. If AF is critical to his needs, then I suspect the D800 might not have been practical for him anyway.

Of course they do, but you still remain determined to argue a losing argument.

I don't see I am losing except you perceive I am

I never said there is no difference but ONLY dispute it's not big when you expose correctly or ETTR in DR and even in resolution when you use some best lenses that is confirmed by DXOMark test.

Again, if you expose correctly and ETTR, the advantages of the D800 will only be widened IMO. And the DXOMark test still gives the D800 a good 6 mpx advantage over the 5D3.

 

[qianp2k]

SubPrime wrote:

qianp2k wrote:

schmegg wrote:

I understand where you're coming from based on who you were replying to initially. ;-)

Or in one year's ago in his tons of D800 vs 5D3 posts in this forum, so I am not surprised he has special passion in his Nikon gear ;-)

I have passion for both ;-)

It's like having two kids who I love equally. I won't hear a bad word said about either one.

Glad to hear. If I had extra money and time (I don't have much free time except weekend spending with family mostly), I'd also build up Nikon system

But saying it is significant because you use both holds no more or less weight than another dual system user who says it isn't significant. Plain fact.

I'd rather to believe well established photographers such as Hans and Galland who also own both systems or used to use both, and both said there is no significant difference. Nobody would suggest them a fanboy in one system or another.

I suspect you would rather to believe the word of someone who agrees with you and makes you feel a little less insecure, in spite of the fact you have no reason to be. The 5D3 is more practical is many cases precisely because it has lower resolution - that doesn't make it less a camera.

No but truly from what I have seen. Yes I don't own a D800/D600 but I have seen enough photos. So extreme shadow pulling capability indeed very impressive that I never actually trying to dispute, but only dispute this is not a good technique in general (doesn't mean in every case) that something I'd shoot even I had a D800 and even my next Canon FF - 6D2 or 5D4 also have 14-stop DR unless it has no obvious side effects. So what I am disputing is not a brand/model specific but in general.

I have already pointed out that Hans is not impractical.

He is the landscape photographer I follow in this forum and his work is very impressive. I actually took his suggestion to shoot in exposure bracket in tricky and contrast scenes so I could select the best one for further processing. BTW serious landscape photographers like him usually don't shoot under noon contrast light time but in early morning or later afternoon golden time so that further not a big deal to them. I, however, just a vacation snapshooter

I am not familiar with Galland's comments.

http://www.dpreview.com/forums/post/50090116

You can chat with him in details. He used to own 5D2. Like some, switched to D700, and then he bought D800. After many months of usage in his pro work, he dislike D800, mainly in AF (before firmware available) and other factors and he didn't see D800 obvious resolution (but he didn't print very big anyway), then he switched to 5D3 and seem happy from there

 

[SubPrime]

In case you do think I am a Nikon fanboy, you might be interested to know that I haven;t picked up any Nikon gear in almost a year and have spent close to 10k this year alone on Canon.

I now won as many Canon lenses as I do Nikon, more speedlights (4 600EX-RTs and SET-E3_Rt controller) and am seriously considering either a second 5D3 body or even a 1DX.;-)

 

[qianp2k]

SubPrime wrote:

In case you do think I am a Nikon fanboy, you might be interested to know that I haven;t picked up any Nikon gear in almost a year and have spent close to 10k this year alone on Canon.

glad to hear.

I now won as many Canon lenses as I do Nikon, more speedlights (4 600EX-RTs and SET-E3_Rt controller) and am seriously considering either a second 5D3 body or even a 1DX.;-)

Great. I wish I could have that luxury to build up dual systems. It's the best to own both and each of them has best is some areas. Canon RT based flash system is now lightly the best in industry. If you can afford 1DX , it's better than 5D3 in every aspect except 4mp shortage. Like 6D 1DX is also less subject to banding if you push a deep underexposed photo.

http://www.dpreview.com/forums/post/50090116

He is the guy I am talking about, Garland

Also I guess you know who is OP

http://xerodigital.ca/canon-1dx-nikon-d4-compare-wedding-photographers/

 

[qianp2k]

SubPrime wrote:

qianp2k wrote:

SubPrime wrote:

qianp2k wrote:

http://www.dpreview.com/forums/thread/3514784

You got kidding me. How many dual system owners we have heard said virtually no difference when print to 30x20" or even 40x30".

You came up with one name. I am a dual system owner who says you can see noticeable difference when print to 30x20"

You must kidding me that if you print to 20x30" from above Rick's samples, you will see noticeable difference?

I wasn't referring to that example.

There are other comparison which reveal that the D800 renders detail the 5D3 cannot even detect and no amount of up-sampling or post processing tricks will change that.

Again I seriously doubt you can see much difference at 30x20" prints if both photos are processed properly. I said that because I heard bunch of dual owners said that and they are all professionals.

As I pointed out to you in another comparison between a D800E and an IQ180, there will be cases when the higher res superior system does not display any noticeable advantage.

the difference is not just 36mp and 80mp but crop format difference. IQ180 has a much bigger sensor that is a dominant factor.

I doubt you will even see noticeable different at 30x40".

You can doubt it, or you can compare actual prints as I have done.

At least not from Rick's samples that I am sure.

I upsampling 5D3 file to match to D800 size and would have pixel peeping to see very small difference. I have a few 20x30" prints from 5D1 and 5D2 and honestly I even don't see much difference either. The difference between D800 and 5D3 is far less than between 5D2 and 5D1 according to DXOMark system test (lens+sensor).

I read bunch of photogeraphers such as Hans Kruse who truly don't have agenda and widely respectied photographer said they don't see much difference. I'd rather believe him than you.

Again, I have pointed out that Kruse was happy to exaggerate the difference between the D8002 and HD-40, which shows even less disparity than the D800 and 5D3 comparisons. Even as the the reviwer, Alex Koloskov, said the difference would not be noticeable on even a large print, Kruse insisted that the differences were significant.

Kruse doesn't personally experience HD-40 I guess but IQ160 that is a different animal and has a noticeable bigger sensor (not all MF cameras have the standard same size sensors unlike FF). I have no clue in Alex. I hope DXOMark can extend tests to many other cameras.

That makes Kruse biased IMO.

You better to check with him. But he seems didn't experience HD-40 but IQ160.

My processed samples are directly out of CS6 by upsampling 5D3 file to D800E size and all others are the same default/Zero setting that everyone can duplicate. Or you processed D800E RAW and I process 5DIII RAw and let's compare?]

To be honest you're strongly Nikon biased from many have seen in your posting history.

I would disagree. If you have looked at my posting history, you will notice that on the Nikon forums, I have been a fierce critic of Nikon, and even been dismissed as a Canon fanboy.

hope not.

And are you seriously going to deny that you are not only strongly Canon biased, but entirely?

I am a fan but not fanboy that would deny everything of Nikon advantages. I never deny Nikon advantages in DR and resolution but just warn overexaggeration.

Sure I only shoot Canon but I have seen enough D800/D600 photos and I don't see they are much better in IQ overall.

But don't you realize how weak that sounds when you are comparing your limited experience with D800/D600 photos to someone who shoots both systems all the time and in all manner of

I heard from many others who actually spent much less time in DPR forums unlike you who heatedly involved in 5D3 and D800 debates then from what I saw ;-)

I have read others who widely regarded as honesy and not brand binded as Galland (may not spell his name correclty). He is 5D2->D700->D800->5D3 shooter and he has true professional bussiness. He also said there is no much difference and he switched back to 5D3 because it has better AF.

I would like to read Galland's comments and find out what kind of shooting he does. If AF is critical to his needs, then I suspect the D800 might not have been practical for him anyway.

http://www.dpreview.com/forums/post/50090116

Search his other posts. He did say several times and seem his opinion is very balanced.

Of course they do, but you still remain determined to argue a losing argument.

I don't see I am losing except you perceive I am

I never said there is no difference but ONLY dispute it's not big when you expose correctly or ETTR in DR and even in resolution when you use some best lenses that is confirmed by DXOMark test.

Again, if you expose correctly and ETTR, the advantages of the D800 will only be widened IMO.

It is not from DR perspective but for resolution perspective that I agree as DPR said you do need to use the best techniques to fully take advantage of D800 resolution but that is true to every camera

And the DXOMark test still gives the D800 a good 6 mpx advantage over the 5D3.

I see less with the same Sigma 35/1.4 lens, one of the most resolvable lenses.

Sigma 35/1.4 on D800 vs 5D3 23 vs 19 mpix

--
http://qianp2k.zenfolio.com/

 

[Phil M Winder]

qianp2k wrote:

Phil M Winder wrote:

qianp2k wrote:

Phil M Winder wrote:

qianp2k wrote:

Phil M Winder wrote:

qianp2k wrote:

Phil M Winder wrote:

qianp2k wrote:

Phil M Winder wrote:

qianp2k wrote:

Marquee wrote:

ETTR means "Expose to the Right", so looking at the histogram, the exposure is as far right as possible without clipping highlights - in reality you allow the highlights to be clipped a tad because the exposure information you get is derived from a Jpeg representation of the RAW you took (assuming you are exposing in a RAW mode), so some clipping can be recovered later. A RAW file that can then take significant pulling of shadow detail could then repsesent a large dynamic range albeit in a form within the viewing limitations of the protraying medium, being monitor or print. Or you could blend two exposures of course which in many cases will be equally good or better (assuming a static subject)

Here are two BIF samples I used ETTR technique. It's under bright early afternoon sunlight at Cancun beach. I was shooting pelican in flight against bright sky and ocean wave. I used Partial metering which is a larger version of Spot metering so already gain +1/3 EV from normal Evaluative metering (that similar to Nikon's Matrix metering).

I gave +2/3 to +1 EC (exposure compensation) on top of Partial metering. So the main subject, the pelicans, were exposed correctly so I don't have to pull shadow aggressively. In OOC RAW files, the highlight, sky or ocean waves were a bit of overblown but I still can recover in LR4 (Recovery bar, software GND filter and Brush tool if necessary). Someone please shows us D800/D600 ETTL (expose to left) BIF photos that exposed on sky then pull birds from deep dark shadows, and let's compare feather details at large size :-D

 

[Phil M Winder]

qianp2k wrote:

Phil M Winder wrote:

You didn't understand that statement, but somehow comparing against Nikon makes sense?

Phil or actually Victor, are you seriously trying to discuss or just want to argue? I don't understand your point so you need to explain. Why you come to here BTW as an exclusive mFT shooter? You even don't own any modern DSLR except very old D60 (as I also own D50).

--
http://qianp2k.zenfolio.com/

I was discussing with you about your perception of ETTR, and how you think it helps in a high contrast, limited DR situation. No argument intended. What you think I own or shoot is irrelevant.

Answer for below:

If I have to elaborate to you about why there is a need to maintain a certain shutter speed at lower light, then you have bigger things to worry about.

You couldn't answer directly about how you determine highlight or shadow limits. A key factor of ETTR or ETTL.

 

[qianp2k]

Phil M Winder wrote:

qianp2k wrote:

Phil M Winder wrote:

qianp2k wrote:

Phil M Winder wrote:

qianp2k wrote:

Phil M Winder wrote:

qianp2k wrote:

Phil M Winder wrote:

qianp2k wrote:

Phil M Winder wrote:

qianp2k wrote:

Marquee wrote:

ETTR means "Expose to the Right", so looking at the histogram, the exposure is as far right as possible without clipping highlights - in reality you allow the highlights to be clipped a tad because the exposure information you get is derived from a Jpeg representation of the RAW you took (assuming you are exposing in a RAW mode), so some clipping can be recovered later. A RAW file that can then take significant pulling of shadow detail could then repsesent a large dynamic range albeit in a form within the viewing limitations of the protraying medium, being monitor or print. Or you could blend two exposures of course which in many cases will be equally good or better (assuming a static subject)

Here are two BIF samples I used ETTR technique. It's under bright early afternoon sunlight at Cancun beach. I was shooting pelican in flight against bright sky and ocean wave. I used Partial metering which is a larger version of Spot metering so already gain +1/3 EV from normal Evaluative metering (that similar to Nikon's Matrix metering).

I gave +2/3 to +1 EC (exposure compensation) on top of Partial metering. So the main subject, the pelicans, were exposed correctly so I don't have to pull shadow aggressively. In OOC RAW files, the highlight, sky or ocean waves were a bit of overblown but I still can recover in LR4 (Recovery bar, software GND filter and Brush tool if necessary). Someone please shows us D800/D600 ETTL (expose to left) BIF photos that exposed on sky then pull birds from deep dark shadows, and let's compare feather details at large size :-D