Pages: 1 ... 4 5 [6] 7 8 9   Go Down

Author Topic: EOS 5Ds- why all this negativity?  (Read 57327 times)

Bart_van_der_Wolf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 8913
Re: EOS 5Ds- why all this negativity?
« Reply #100 on: February 10, 2015, 12:31:28 pm »

I think the term negativity was a poor choice in the original question.  It is more disappointment than negativity.  Everyone already knew the MP count and were hoping to get DR and maybe a lower base ISO out of the deal as well.

Yes, disappointment would be more appropriate. However, I still think that most people do not understand the positive MTF effect of a denser sampling of the projected image. All lenses get instantly better, with a higher MTF modulation.

Cheers,
Bart
Logged
== If you do what you did, you'll get what you got. ==

shadowblade

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2839
Re: EOS 5Ds- why all this negativity?
« Reply #101 on: February 10, 2015, 04:26:28 pm »

Yes, disappointment would be more appropriate. However, I still think that most people do not understand the positive MTF effect of a denser sampling of the projected image. All lenses get instantly better, with a higher MTF modulation.

Cheers,
Bart

I think the disappointment is more with the claimed dynamic range than the resolution. Obviously, it sets a new standard in resolution, but Sony/Nikon are likely to match that within months. But Canon set a new standard in IQ with the 1Ds3/5D2 - and left us there for the next seven years. And now, it seems, for another three or four at least. A real disappointment for those of us who bought into the system from MF film after the 5D2 was released.
Logged

BernardLanguillier

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 13983
    • http://www.flickr.com/photos/bernardlanguillier/sets/
Re: EOS 5Ds- why all this negativity?
« Reply #102 on: February 10, 2015, 05:21:13 pm »

I think the disappointment is more with the claimed dynamic range than the resolution. Obviously, it sets a new standard in resolution, but Sony/Nikon are likely to match that within months. But Canon set a new standard in IQ with the 1Ds3/5D2 - and left us there for the next seven years.

The high DR era indeed started with the D3x on 1-Dec-2008, which came out one+ year after the 1Ds3 and only a few months after the 5D2. The D3x already had nearly 2 stops more DR (and still 2 stops more DR compared to a 5DIII).

http://www.dxomark.com/Cameras/Compare/Side-by-side/Nikon-D3X-versus-Canon-EOS-5D-Mark-II-versus-Canon-EOS-1Ds-Mark-III___485_483_436

The D800/D810 improved this another half/full stop, but the paradigm shift with real impact on shooting practise happened with the D3x.

I would relax until actual measurements tell us where the 5Ds stands. Most of this agitation is based on some clumsy comments by some Canon USA marketing guys. They may have gotten it wrong.

Cheers,
Bernard
« Last Edit: February 10, 2015, 05:44:11 pm by BernardLanguillier »
Logged

shadowblade

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2839
Re: EOS 5Ds- why all this negativity?
« Reply #103 on: February 10, 2015, 09:03:13 pm »

The high DR era indeed started with the D3x on 1-Dec-2008, which came out one+ year after the 1Ds3 and only a few months after the 5D2. The D3x already had nearly 2 stops more DR (and still 2 stops more DR compared to a 5DIII).

http://www.dxomark.com/Cameras/Compare/Side-by-side/Nikon-D3X-versus-Canon-EOS-5D-Mark-II-versus-Canon-EOS-1Ds-Mark-III___485_483_436

The D800/D810 improved this another half/full stop, but the paradigm shift with real impact on shooting practise happened with the D3x.

I would relax until actual measurements tell us where the 5Ds stands. Most of this agitation is based on some clumsy comments by some Canon USA marketing guys. They may have gotten it wrong.

Cheers,
Bernard

Yep.

Although the D3x cost $8k, was useless at more than ISO 400 or so and only shot at slightly over a frame per second, so it wasn't much of an all-rounder. The D800/D800e changed all that.
Logged

Ghibby

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 72
Re: EOS 5Ds- why all this negativity?
« Reply #104 on: February 10, 2015, 09:21:29 pm »

Just seen this over on Northlight, suggesting seriously improved DR and colour accuracy compared to any existing Canon camera we have seen.

Quote
Several more people with testing experience of the new 5Ds have written to agree with some of the comments in the previous days.
In particular, one commented that:
"Canon's new 50.6mp sensor at low ISO will perform much better then any other EOS camera currently in the pro line.
The colour filters on the sensor are designed to produce a higher level of colour accuracy and separation, the sensor itself runs at a significantly lower temperature.
This will come at a price since high ISO performance will drop significantly.
The new sensor will capture 14 stops of DR (just like the 5D III) ... However it will produce remarkably cleaner results when lifting deep shadows".
Once again this is not something I can directly verify.

http://www.northlight-images.co.uk/cameras/Canon_5ds.html

The fly in the ointment in this quote is suggesting that the 5D3 is capable of 14 stops DR which it clearly is not.

Hopefully the photographer-centric design approach Canon have taken with regard to mirror vibration and body stiffness and so on have also been applied to the actual sensor architecture of this new camera.

Ben
Logged

shadowblade

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2839
Re: EOS 5Ds- why all this negativity?
« Reply #105 on: February 10, 2015, 09:41:23 pm »

Just seen this over on Northlight, suggesting seriously improved DR and colour accuracy compared to any existing Canon camera we have seen.

http://www.northlight-images.co.uk/cameras/Canon_5ds.html

The fly in the ointment in this quote is suggesting that the 5D3 is capable of 14 stops DR which it clearly is not.

Hopefully the photographer-centric design approach Canon have taken with regard to mirror vibration and body stiffness and so on have also been applied to the actual sensor architecture of this new camera.

Ben

I can only hope that '14 stops of DR' is correct and not 'like the 5D3'. Hopefully, when the Canon reps (who are obviously marketing/sales people rather than technical/engineering people) mentioned that DR was 'equal' to the 5D3, they meant 'no worse than' the 5D3 (i.e. better) rather than 'the same as' the 5D3... people use the term 'equal' in a number of ways, and not always in the scientific or mathematical sense.
Logged

Geraldo Garcia

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 470
    • Personal blog
Re: EOS 5Ds- why all this negativity?
« Reply #106 on: February 10, 2015, 10:10:47 pm »

I can only hope that '14 stops of DR' is correct and not 'like the 5D3'. Hopefully, when the Canon reps (who are obviously marketing/sales people rather than technical/engineering people) mentioned that DR was 'equal' to the 5D3, they meant 'no worse than' the 5D3 (i.e. better) rather than 'the same as' the 5D3... people use the term 'equal' in a number of ways, and not always in the scientific or mathematical sense.

There is another mention on Keith's page, supposedly from an actual tester of the 5dsR, that says: "Low ISO DR is put at 1.5-2 stops better than the 1D X, but high ISO performance (6400) falls marginally behind the 7Dmk2".

Hummm... But if that is true, why Canon isn't using it as marketing?
Logged

Telecaster

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3686
Re: EOS 5Ds- why all this negativity?
« Reply #107 on: February 10, 2015, 11:35:51 pm »

Personally I prefer to hold back on judging any photo product until real-world users are putting it to real-world use. I particularly don't pay much attention to PR flacks or beta testers (unless I happen to know the beta testers, in which case private discussions typically reveal more nuanced views than do public statements).

-Dave-
Logged

shadowblade

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2839
Re: EOS 5Ds- why all this negativity?
« Reply #108 on: February 11, 2015, 12:12:36 am »

Personally I prefer to hold back on judging any photo product until real-world users are putting it to real-world use. I particularly don't pay much attention to PR flacks or beta testers (unless I happen to know the beta testers, in which case private discussions typically reveal more nuanced views than do public statements).

-Dave-

I'm generally more interested in formal, objective test results than real-world use and, especially, sample photos. This is because sample photos are very user-dependent and don't tell me anything about the limitations and capabilities of the camera, only how good the photographer was. And opinions on real-world use don't say very much either, since most real-world use (by definition) stays well within the technical capabilities of the equipment. As a potential user of any equipment, I'm interested in defining the performance limits of equipment - ISO, noise, DR - so that I know whether they actually expand the boundaries of what i'm able to do with it.

Same goes with everything else I buy, really, be it cars, medical equipment or power tools. Objective measurements beat a subjective review any day.
Logged

Bart_van_der_Wolf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 8913
Re: EOS 5Ds- why all this negativity?
« Reply #109 on: February 11, 2015, 04:03:36 am »

There is another mention on Keith's page, supposedly from an actual tester of the 5dsR, that says: "Low ISO DR is put at 1.5-2 stops better than the 1D X, but high ISO performance (6400) falls marginally behind the 7Dmk2".

Hi Geraldo,

I take those comments as subjective, they are just impressions. But I put them in the same category of previous claims about the 16 stop DR of medium format Digital backs, not substantiated by cold hard facts.

Quote
Hummm... But if that is true, why Canon isn't using it as marketing?

Probably because Canon knows it isn't so. What may be true, but we'll have to wait and see, is that the pattern noise is reduced. That would be helpful, because the remaining read-noise + low shot-noise is more random (easier to remove) and less obnoxious. It allows better demosaicing as well, if one doesn't have to hold back in order to avoid amplifying pattern noise.

The way I read the comments so far, is that it is mostly due to the help of the dual Digic6 performance that the data can be cleaned up a bit, but that the inherent CMOS electronics have not been improved enough to really lower the read noise like on the Sony sensors. Which is already something if you account for the much smaller sensels.

So read noise is probably about the same, pattern noise is reduced, and the photographic S/N ratio (not that of the sensor in isolation) is improved due to denser sampling of the lens projected image. There may be other improvements in lower noise generating components, but that would be unsubstantiated speculation. Let's wait for the DxO analysis on noise and DR (I expect 11.5 stops or slightly more), or for people to get their hands on some Raws to analyze.

I expect an overall cleaner image, with higher resolution, and more robust data for post-processing due to the improved photographic S/N ratio. Let's also not forget the role of the Raw converter. Some converters are consistently producing technically better, more easy to work with raw conversions, and others do all sort of funky tone compression stuff that first needs to be tamed, if the aliasing and mazing artifacts are even not creating another challenge. Capture One offers a very nice low artifact conversion, with lots of control that is not swamped with under the hood automatic tone adjustments (unless the user specifically tells it to have a go).

I recently had a look at Canon DPP's Digital Lens Optimization (DLO) for images I shot with the older EF 16-35mm f/2.8, and the DLO even improved the corners towards being useful. The DLO creates and embeds an additional corrected Raw instead of just postprocessing the Raw converted data, which kind of proves that Raw converters (working on better Raw data) and intimate knowledge/understanding of the Raw data can make a huge difference. Luckily Canon has (so far) not taken the route to actively reduce noise or compress tone-curves by default, before writing the Raw data. The data quality is usually as Raw as Raw can be (including an offset to not cripple the read noise), which offers an excellent starting point for those who know how to do a proper Raw conversion.

Cheers,
Bart
Logged
== If you do what you did, you'll get what you got. ==

Martin Wouterlood

  • Newbie
  • *
  • Offline Offline
  • Posts: 47
Re: EOS 5Ds- why all this negativity?
« Reply #110 on: February 11, 2015, 12:44:55 pm »

This new Canon EOS 5DS / 5DSR (why not just 5DR?) pair may not be heralded as the new ant's pants in dynamic range, but for me it's possibly good enough...
or the bee's knees in high iso, but also quite adequate...
or the duck's n*ts in under expose / over in processing to turn black into daylight, but again...(you got it)
but if the R & D energy put into this thing really means it can (with disciplined technique) deliver some of the best detail, accurate colour, high and low light tolerance currently available, straight out of the can,
then it will finally be showing some real pedigree and Canon will get my money.


Logged

Phil Indeblanc

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2017
Re: EOS 5Ds- why all this negativity?
« Reply #111 on: February 11, 2015, 01:19:44 pm »

This new Canon EOS 5DS / 5DSR (why not just 5DR?) pair may not be heralded as the new ant's pants in dynamic range, but for me it's possibly good enough...
or the bee's knees in high iso, but also quite adequate...
or the duck's n*ts in under expose / over in processing to turn black into daylight, but again...(you got it)
but if the R & D energy put into this thing really means it can (with disciplined technique) deliver some of the best detail, accurate colour, high and low light tolerance currently available, straight out of the can,
then it will finally be showing some real pedigree and Canon will get my money.

+1
Logged
If you buy a camera, you're a photographer...

Telecaster

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3686
Re: EOS 5Ds- why all this negativity?
« Reply #112 on: February 11, 2015, 01:42:58 pm »

I'm generally more interested in formal, objective test results than real-world use and, especially, sample photos.

Well, I wasn't referring to Jane & Joe Happysnapper.  :D  In my experience serious owners/users tend to put high-end cameras like these through the ringer. Issues often come up that beta testers either miss or at least don't publicly report. And…I care about a lot more than just image quality. Handling & feel, how clunky or fluid the UI is, etc. Thus my preference for holding back judgment 'til folks with skin in the game have their say.

Edit: interesting how dwswager (just below) & I both use the phrase skin in the game in our posts. I think he's referring to the brand affinity beta testers often (okay, pretty much always) feel, which can get in the way of objectively evaluating a new product. (Given that pure objectivity is impossible.) Whereas I'm referring to the kind of person who lays down her/his money and then has a heightened interest in the damn camera measuring up to the outlay.  ;)  And who will be vocal about it if it doesn't.

-Dave-
« Last Edit: February 11, 2015, 03:01:57 pm by Telecaster »
Logged

dwswager

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1375
Re: EOS 5Ds- why all this negativity?
« Reply #113 on: February 11, 2015, 02:12:19 pm »

Personally I prefer to hold back on judging any photo product until real-world users are putting it to real-world use. I particularly don't pay much attention to PR flacks or beta testers (unless I happen to know the beta testers, in which case private discussions typically reveal more nuanced views than do public statements).

-Dave-

Exactly!  Until a product like this gets into the hands of a significant number of real user's hands, you have little idea what the actual performance is going to be.  Beta testers typically have skin in the game already.  Most aren't intentionally misleading, but the bias is baked in.  You get used to the quirks and issues in developmental products and then assume they are supposed to be there.

Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Re: EOS 5Ds- why all this negativity?
« Reply #114 on: February 15, 2015, 06:15:17 pm »

Fortunately, when I switched to Nikon about 3 years ago, I kept all my Canon lenses. The last Canon upgrade I made was the purchase of the 50D. Whilst it's impossible to be certain about the extent of improvement of DR in the new camera, until the model is tested, there has already been a precedent set by Nikon when they introduce a 36mp D800 that had essentially the same pixels as the earlier, cropped-format D7000.

If we check the test results at DXOMark, comparing the Nikon D7000 with the D800 or D800E, we see that at a given print size, the SNR of the D800 is about 4-5 dB greater than that of the D7000 and DR is around one full stop greater.

Since Canon representatives have claimed the 5DS pixel is essentially the same as the 7D2 pixel, we can make a reasonable deduction about the DR and SNR of the 5DS by checking the results for the 7D2 at the DXOMark website.

According to DXOMark tests, the DR of the 7D Mk II, between ISO 100-400, is equal to that of the 5D Mk III and the 1Ds Mk III. That's a significant improvement for a cropped format to have the same DR as a full-frame.

It is therefore reasonable to expect that the 5DS will have about one stop better DR than the 5D3 or 1Ds3, between ISO 100-400. If the shadows are also free of banding and pattern noise, as the 7D2 is, then that's also an additional, worthwhile improvement.

The significant increase in the resolution of the sensor to a massive 50mp effectively upgrades all of one's Canon lenses in one fell swoop. Even if you think you don't need 50mp, it's still an advantage. A 50mp image shot with a zoom lens, when down-sampled to say 23mp, will look like a shot from the 5D3 using a prime lens. A 50mp image shot with a merely good quality prime, when down-sampled to 23mp, will look like a shot from the 5D3 using a superb or excellent quality prime.

I'm not sure I'll be able to resist buying this new model.  ;)
Logged

Rory

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 528
    • Recent images
Re: EOS 5Ds- why all this negativity?
« Reply #115 on: February 15, 2015, 07:32:41 pm »

It is therefore reasonable to expect that the 5DS will have about one stop better DR than the 5D3 or 1Ds3, between ISO 100-400. If the shadows are also free of banding and pattern noise, as the 7D2 is, then that's also an additional, worthwhile improvement.

I share your optimism about the SNR but I have doubts about your DR conclusions.  I'm betting it will be 11.8 Evs compared to the 14.8 for the D810.  I'll go out on a limb and guess the DxO score for low-light ISO is in the 2600-2800 range compared to the 7DMKII 1082.  It will still be a fine camera and I'm also hoping for stellar colour rendition.
Logged
[url=http://www.flickr.com/photos/roryhi

shadowblade

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2839
Re: EOS 5Ds- why all this negativity?
« Reply #116 on: February 15, 2015, 07:56:48 pm »

I share your optimism about the SNR but I have doubts about your DR conclusions.  I'm betting it will be 11.8 Evs compared to the 14.8 for the D810.  I'll go out on a limb and guess the DxO score for low-light ISO is in the 2600-2800 range compared to the 7DMKII 1082.  It will still be a fine camera and I'm also hoping for stellar colour rendition.

Even if they just scaled up the 7D2 and kept the same pixel quality, the standardised DxOMark DR would be 11.8 (the 7D2's pixel quality) + 1.32 (log base 2 of 50.6/20.16) = 13.12, based on the resolution alone.

To achieve a DR of only 11.8, they'd have to put in really crummy pixels...
Logged

Rory

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 528
    • Recent images
Re: EOS 5Ds- why all this negativity?
« Reply #117 on: February 15, 2015, 08:13:29 pm »

Even if they just scaled up the 7D2 and kept the same pixel quality, the standardised DxOMark DR would be 11.8 (the 7D2's pixel quality) + 1.32 (log base 2 of 50.6/20.16) = 13.12, based on the resolution alone.

To achieve a DR of only 11.8, they'd have to put in really crummy pixels...

You are assuming DR scales like SNR.  What is the basis for that assumption?
Logged
[url=http://www.flickr.com/photos/roryhi

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Re: EOS 5Ds- why all this negativity?
« Reply #118 on: February 15, 2015, 09:12:34 pm »

You are assuming DR scales like SNR.  What is the basis for that assumption?

DR is basically SNR at low signal levels. The composition of the noise at low signal levels might be different to the noise at 18% grey, but the nature of the larger sensor, which gathers more light from the same scene, is that noise across the whole range will be reduced.

The D5S sensor has approximately 2.5x the area of the 7D2 sensor and therefore gathers about 2.5x as much light, which is more than an additional stop. One would expect noise to be reduced by at least a stop across the visible range. Of course, noise is more of a problem in the shadows and the mid-tones, which is why DXO produce test results for DR and SNR at 18% grey which is about the level of skin tones.

My impression, comparing camera sensors at DXOMark, is that the DR of the 5Ds will be improved by approximately 1 stop (or 1 EV), compared with the 7D2, and SNR at 18% will be improved by 4-5dB, which is a bit more than a stop. This is assuming the 5DS pixel and associated electronics is of the same quality as the 7D2 pixel.
Logged

Rory

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 528
    • Recent images
Re: EOS 5Ds- why all this negativity?
« Reply #119 on: February 15, 2015, 09:17:38 pm »

I guess we will have to wait and see.
Logged
[url=http://www.flickr.com/photos/roryhi
Pages: 1 ... 4 5 [6] 7 8 9   Go Up