Pages: 1 ... 6 7 [8] 9 10 11   Go Down

Author Topic: I would like to understand the MF look.  (Read 69441 times)

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: I would like to understand the MF look.
« Reply #140 on: December 09, 2013, 02:11:53 pm »

Hi,

I have great respect for BC's images, but it is not what I would typically shoot. More something like these (lots of different cameras, APS-C to MFD):

P45+

Sony Alpha 700?

Sony Alpha 700

Sony Alpha 900

Sony Alpha 900

Sony Alpha 77


Best regards
Erik


Hi Sheldon,

I agree. And what's more, a photon looks the same to CCD or CMOS. Most of the alleged differences can be applied to any image in post-processing, in particular color differences depend heavily on the profile used. The examples shown also exhibit lots of dedicated lighting setups, which makes a difference regardless of the camera platform.

Cheers,
Bart
« Last Edit: December 10, 2013, 12:42:04 am by ErikKaffehr »
Logged
Erik Kaffehr
 

bcooter

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1520
Re: I would like to understand the MF look.
« Reply #141 on: December 09, 2013, 05:57:02 pm »

Hi Sheldon,

I agree. And what's more, a photon looks the same to CCD or CMOS. Most of the alleged differences can be applied to any image in post-processing, in particular color differences depend heavily on the profile used. The examples shown also exhibit lots of dedicated lighting setups, which makes a difference regardless of the camera platform.

Cheers,
Bart

I don't even know what a photon is and I'm not being flip about anyone that does know what a photon is. 

But the "alleged" differences are real to me and I don't see a difference every now and then I see a difference every time I use medium format, which is not as often as I'd like.

The camera doesn't make the photographer, but the camera can enhance the photograph or make it more difficult.

I personally see a place for medium format and yes I think there is a drastic difference between ccd capture and cmos.

IMO

BC
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
Higher priority on color accuracy is likely the cause of the CCD/CMOS difference
« Reply #142 on: December 09, 2013, 06:46:08 pm »

But the "alleged" differences are real to me and I don't see a difference every now and then I see a difference every time I use medium format, which is not as often as I'd like.
...
I think there is a drastic difference between ccd capture and cmos.
I will not dispute that you see a systematic difference between:
(a) CCD sensors from Kodak/Truesense and Teledyne-Dalsa, mostly in formats larger than 36x24mm
and
(b) CMOS sensors from Sony, Canon, etc. in formats 36x24mm and smaller.

But there are several systematic differences between these two classes of sensors, and  of these, the difference in how the signal is handled by the sensor _after_ it has arrived and been converted from light to an electric charge (which is the only significant difference between CCD and CMOS) is almost certainly not the difference that leads to a different look.

The likely cause is instead different design decisions for the color filter arrays, where the MF sensors seem to put a greater priority on color accuracy at the expense of lower sensitivity and thus worse low-light performance, whereas the CMOS sensors are mostly used on cameras where competitive pressure is more often based on comparisons of low-light performance, pushing the designs to skew towards better sensitivity at the cost of worse color accuracy.

By the way, this trade-off happens because one way to increase sensitivity ("quantum efficiency") is to have the color filters let through a wider range of colors, so that instead of being R or G or B, they are more like R+g, r+G+b, and g+B.

This gets worse as you go down the sensor size scale: the sensors in most compact cameras have suspiciously high quantum efficiency compared to SLR-sized sensors, a sign that their CFAs are letting through a broader range of colors.
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto

Bill,

I share your analysis to a part. Interestingly BC also preferred Leica M9 rendition to Leica M(240).

Another interesting observation is that the images BC posted are taken at relative high ISO, and he mentions that he prefers the slightly grainy look of the DBs. He says that the MFD images are not so smooth as the DSLR images.

I have not used my P45+ above 50 ISO, so I don't know. I also don't shoot talent in artificial light, much because of lack of talent.

Best regards
Erik

I will not dispute that you see a systematic difference between:
(a) CCD sensors from Kodak/Truesense and Teledyne-Dalsa, mostly in formats larger than 36x24mm
and
(b) CMOS sensors from Sony, Canon, etc. in formats 36x24mm and smaller.

But there are several systematic differences between these two classes of sensors, and  of these, the difference in how the signal is handled by the sensor _after_ it has arrived and been converted from light to an electric charge (which is the only significant difference between CCD and CMOS) is almost certainly not the difference that leads to a different look.

The likely cause is instead different design decisions for the color filter arrays, where the MF sensors seem to put a greater priority on color accuracy at the expense of lower sensitivity and thus worse low-light performance, whereas the CMOS sensors are mostly used on cameras where competitive pressure is more often based on comparisons of low-light performance, pushing the designs to skew towards better sensitivity at the cost of worse color accuracy.

By the way, this trade-off happens because one way to increase sensitivity ("quantum efficiency") is to have the color filters let through a wider range of colors, so that instead of being R or G or B, they are more like R+g, r+G+b, and g+B.

This gets worse as you go down the sensor size scale: the sensors in most compact cameras have suspiciously high quantum efficiency compared to SLR-sized sensors, a sign that their CFAs are letting through a broader range of colors.
Logged
Erik Kaffehr
 

synn

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1235
    • My fine art portfolio
Re: I would like to understand the MF look.
« Reply #144 on: December 09, 2013, 08:37:54 pm »

Bcooter, what is this sorcery? You are using your digital backs to make actual photographs you say? You are basing your opinions on years of use in a professional setting you say? There are more important things to a photograph than what can be quantified in numbers you say? BLASPHEMY I SAY! Now repent by deleting your posts and let this forum go back to it's regular broadcast of internet experts posting images of dollar bills and brick walls, numbers and graphs. Lest you be burned at the pixelpeeper-stake for heresy!

All hail the Canikon Dx900XDs1, hallowed be thy name.

/end sarcasm

Haha.

It really is refreshing to see real images instead of charts and graphs, innit?
BC's images have really helped me make my decision to go MF. Can't say the same about all the numerical analysis.
Logged
my portfolio: www.sandeepmurali.com

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: I would like to understand the MF look. Another look
« Reply #145 on: December 09, 2013, 09:16:17 pm »

Hi,

This topic led me to look trough a lot of images. I normally present my images as slide shows in 1080P format, that gets rid of the resolution difference, and I cannot really say which is which.

Going back to the images I found that there are quite a few images where the MFD stuff was better. What I also found that when I was shooting in with MFD and DSLR in parallell I got a lot of different images. Here are some images that were shot with DSLR (mostly freehand) and MFD (on tripod) essentially standing at the same spot, close to the tripod. All images are also available as raw files. Processing is quick and dirty, in part.

http://echophoto.dnsalias.net/ekr/Articles/MFDJourney/RawImages/OneSpot/

An observation is that using primes on the P45+ forces certain compositions which may turn out better than what I would get with DSLR and zoom lens. What I also have seen that I stitch more with MFD, an image is often just a bit to wide for the lens I use and I don't want to switch to a wider one, so I just make 2/3 exposures and stitch.

Best regards
Erik
« Last Edit: December 09, 2013, 09:39:02 pm by ErikKaffehr »
Logged
Erik Kaffehr
 

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600

I share your analysis to a part. Interestingly BC also preferred Leica M9 rendition to Leica M(240).
The M9 is the exception that proves the rule: the M9 is 36x24mm format, but uses a Kodak CCD with Kodak's color filter designs (at least, I am fairly sure that Leica did not tell Kodak to modify its CFAs in a way that sacrifices color accuracy for the sake of better low-light performance). On the other hand, I do not know anything about the CFA design that CMOSIS uses in its CMOS sensor for the M(240).
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto

Hi,

I don't think it is about colour accuracy, it is more about pleasant colour.

This article by Tim Parkin may be worth checking out: https://www.onlandscape.co.uk/2012/02/the-myth-of-universal-colour/

Best regards
Erik

The M9 is the exception that proves the rule: the M9 is 36x24mm format, but uses a Kodak CCD with Kodak's color filter designs (at least, I am fairly sure that Leica did not tell Kodak to modify its CFAs in a way that sacrifices color accuracy for the sake of better low-light performance). On the other hand, I do not know anything about the CFA design that CMOSIS uses in its CMOS sensor for the M(240).
Logged
Erik Kaffehr
 

Sheldon N

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 828
Re: I would like to understand the MF look.
« Reply #148 on: December 09, 2013, 11:41:14 pm »

If we're sharing pictures, all I've got is some snaps of my kids.... Three are taken with an Aptus 22 RZ67 and three are Canon.











Logged
Sheldon Nalos
[url=http://www.flickr.com

EricWHiss

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2639
    • Rolleiflex USA
Re: I would like to understand the MF look.
« Reply #149 on: December 10, 2013, 12:18:54 am »

Sheldon,
Some really nice work there on your flickr page!
Eric
Logged
Rolleiflex USA

Sheldon N

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 828
Re: I would like to understand the MF look.
« Reply #150 on: December 10, 2013, 12:39:42 am »

Thanks, Eric.
Logged
Sheldon Nalos
[url=http://www.flickr.com

bcooter

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1520

I

But there are several systematic differences between these two classes of sensors,..................snip

I dunno for you your right for me, not.  I also drop a lot of files into a non linear editor and can usually see a big difference, to the point I have to add a layered tint on a medium format file and make it more globlal in color and also add slight lens blur to match a cmos file.

But my images aren't the best examples.  Sheldon's are.

Up close on the tight head shot of the little girl, it looks like any camera.  I have a beauty shot on a computer right now we're working that is from a 1ds3 and I can't tell a huge difference, but if I pulled back like Sheldon did on the little girl with the green outfit it shows and to me it's dramatic.  It also shoes on the boy on the curb.  CMOS does not render that way.

In regards to a m8, m9 or a medium format back, given the same sensor, (ccd) the bigger the better, but even at 35mm size to me the M8 and M9 are far superior to a m240.  I love leicas and if you give me a 240 I'll sell it to KEH.  Give me an M9 I'll keep it, give me an S and I'll kiss you on the lips (I can afford an S the kiss is just for fun*).

IMO

BC

*I'm very secure in my masculinity, not that there is anything wrong with that if I wasn't.  (I think I covered that).
« Last Edit: December 10, 2013, 12:44:04 am by bcooter »
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto

Hi,

Indeed fine images on Sheldon's site. But you may have missed the fine print ;-)

Best regards
Erik
I dunno for you your right for me, not.  I also drop a lot of files into a non linear editor and can usually see a big difference, to the point I have to add a layered tint on a medium format file and make it more globlal in color and also add slight lens blur to match a cmos file.

But my images aren't the best examples.  Sheldon's are.

Up close on the tight head shot of the little girl, it looks like any camera.  I have a beauty shot on a computer right now we're working that is from a 1ds3 and I can't tell a huge difference, but if I pulled back like Sheldon did on the little girl with the green outfit it shows and to me it's dramatic.  It also shoes on the boy on the curb.  CMOS does not render that way.

In regards to a m8, m9 or a medium format back, given the same sensor, (ccd) the bigger the better, but even at 35mm size to me the M8 and M9 are far superior to a m240.  I love leicas and if you give me a 240 I'll sell it to KEH.  Give me an M9 I'll keep it, give me an S and I'll kiss you on the lips (I can afford an S the kiss is just for fun*).

IMO

BC

*I'm very secure in my masculinity, not that there is anything wrong with that if I wasn't.  (I think I covered that).
Logged
Erik Kaffehr
 

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: I would like to understand the MF look.
« Reply #153 on: December 10, 2013, 01:10:23 am »

Hi,

Eric is right! Great images!

Best regards
Erik

Thanks, Eric.
Logged
Erik Kaffehr
 

bcooter

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1520

Hi,

Indeed fine images on Sheldon's site. But you may have missed the fine print ;-)

Best regards
Erik

See.  I think all photographs that are pretty MUST be from a ccd camera.

Maybe ol' Sheldon is slippin one by us.

BC
Logged

Sheldon N

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 828
Re: I would like to understand the MF look.
« Reply #155 on: December 10, 2013, 01:55:08 am »

Haha! Too funny guys.

Erik, that's cheating going to check out the cross posting from my Flickr. :)

Tight headshot of my daughter is the Canon 1DX with 85 f/1.2 at f/2 with a 12mm extension tube to get up real close. The 85 f/1.2 gets all dreamy with an extension tube, aberrations... but in a good way.

My son on the curb is Aptus 22 with RZ67, the 150mm f/3.5 lens wide open. I love that lens, and I agree it's definitely got the MF look.

My daughter in the fairy princess costume is the 1DX and 50mm f/1.2 at f/1.4. In bcooter's defense, that one has been worked over in post a little with selective sharpening.
« Last Edit: December 11, 2013, 11:11:09 am by Sheldon N »
Logged
Sheldon Nalos
[url=http://www.flickr.com

jerome_m

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 670
Re: I would like to understand the MF look.
« Reply #156 on: December 10, 2013, 09:13:21 am »

Haha! Too funny guys.

Erik, that's cheating going to check out the cross posting from my Flickr. :)

Tight headshot of my daughter is the Canon 1DX with 85 f/1.2 at f/2 with a 12mm extension tube to get up real close. The 85 f/1.2 gets all dreamy with an extension tube, aberrations... but in a good way.

My son on the curb is Aptus 22 with RZ67, the 150mm f/3.5 lens wide open. I love that lens, and I agree it's definitely got the MF look.

My daughter in the fairy princess costume is the 1DX and 50mm f/1.2 at f/1.4. In bcooter's defense, that one has been worked over in post a little with selective sharpening.

So, members on this forum failed to recognize the "MF look" when presented with your 6 pictures. I think that sums up the utility of this thread.
Logged

hjulenissen

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2051
Re: I would like to understand the MF look.
« Reply #157 on: December 10, 2013, 09:35:22 am »

I think it is kind of strange if anyone claims that "camera A produce better images than camera B because it has sensor technology X", then proudly proclaim that they could care less about technology? Is inconsequential the right word (English not being my native language).

I mean, it is perfectly kosher to say that I prefer the images that I have been able to get out of camera A to those I have been able to get out of camera B (for whatever reason). Without getting into a technology debate about the hows and whys.

When you start claiming that CCD has inherent technology advantages/disadvantages vs CMOS, then you are making a theoretic, technical statement (possibly based on sensible observations, but seemingly not a scientific systematic approach to observing differences). That is leaving the subjective/artistic world and moving into the objective/scientific world, and you should not be surprised to see counter-arguments based on (ones understanding of) physics or science.

An obvious critique is that of correlation vs causality. Perhaps images that was produced on CCD cameras and presented on this forum are deemed, on average, "1 subjective point" better vs images that was produced on CMOS cameras and presented on this forum. Possible reasons for this (if true):
1. CCD has inherent benefits
2. Cameras using CCDs tends to have advantages over cameras using CMOS (that is not directly caused by sensor choice)
3. People using CCD cameras tends to post only their best images to this forum, while CMOS users posts some moderately good ones as well.
4. CCD cameras have existed for a longer time than CMOS cameras, thus, the accumulated number of amazing images might be expected to be higher for CCD
5. People using CCD cameras are better photographers than people using CMOS cameras
6. People judging images tends to be more positive towards images that they know to be produced by certain camera (technology)

I would suggest that it is impossible to be certain if any of the above are the most important factors (or if the most important factor was left out). We can certainly speculate.

I am frequently frustrated by side-by-sides where someone obviously tried to make each camera shine on its own terms, using whatever development/photoshop trick he/she knew. While this may seem like a good idea, what you get is two subjectively different images and (for my part) none the wiser. I'd rather see one (or several) interpretations where a skilled photoshop user tried to make one of the files look as good as possible/as realistic as possible/as close as possible to the photographers intention, then "force" the other file to be as similar as possible (including noise/loss of details in shadows, posterization, if applicable). An estimate of the time spent and number/complexity of operations would be enlightening.

An other alternative would be to use testcharts in the scene for recording color/blur/... and use something like RawTherapee to make matching "sort-of-objective but still pleasing" output using documented algorithms. The necessary difference in sharpening, noise reduction, color correction matrix etc would be interesting to correlate with the visual output.

-h
« Last Edit: December 10, 2013, 09:55:26 am by hjulenissen »
Logged

TMARK

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1841
Re: I would like to understand the MF look.
« Reply #158 on: December 10, 2013, 09:47:00 am »

So True.  M9 out of the camera with basic adjustments is fantastic.  I usually crush the blacks a bit and desat, no sharpening.  Looks great, much like a P30+ file.  Dare I say Kodachrome-esque.

If you want pretty straight out of camera go M9. If you want a more flexible file go M240.
Logged

jerome_m

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 670
Re: I would like to understand the MF look.
« Reply #159 on: December 10, 2013, 10:00:59 am »

I think it is kind of strange if anyone claims that "camera A produce better images than camera B because it has sensor technology X", then proudly proclaim that they could care less about technology? Is inconsequential the right word (English not being my native language).

The first think that should be checked is: can they tell the cameras apart in a blind test?
Logged
Pages: 1 ... 6 7 [8] 9 10 11   Go Up