August 01, 2014, 06:05:05 AM

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - dilbert

Pages: 1 ... 105 106 [107] 108 109 ... 181
1591
EOS Bodies / Re: Will there be a 7d2 at all? - No
« on: June 11, 2012, 02:06:12 AM »
Missing option: there will be no 7D Mark II at all.

1592

(portrait color depth bits) + (landscape dynamic range) + sqrt(sports low light ISO)
24                                                         11                                47
= 82


I don't think sqrt(low light ISO) makes sense -- you need to take the log of it. The other two scales are in stops, so the third score needs to be transformed to get it into stops. You might need to transform it to get it right after that (e.g. add an offset and a multiplier).

The way you're doing it with sqrt instead of log, the function will grow too quickly (which is why you get lower scores for small sensor cameras)

I tried log but then those three numbers alone did not work (the numbers were too small.)

But it could be that it is "portrait+landscape+log(lowlight)+something_else".

1593
So as far as I'm concerned, DxO's dynamic range numbers are absolutely valid.

I agree.  Allow me to clarify.  I see no issues with DxOMark's measurements - they are useful and accurate, and they disclose their methods to generate those measurements. (FWIW, I personally prefer and use DxO Optics Pro as a RAW converter.)

The problem is the DxOMark Sensor Score.  Here's what Luminous Landscape states:

Documentation about the way the final DxOMark Sensor score is computed from Dynamic Range, Color Sensitivity and Low-light ISO scores is not currently available.

That same article (linked in a thread started by dilbert) indicates there is some 'non-linearity' in the way the Score is calculated.  So...a weighted average, but undisclosed weighting factors, no confirmation that those unknown factors are even consistently applied, non-linear calculations, etc., to me, that makes the DxOMark Sensor Score a meaningless and useless number (not measurement, just an irrelevant number).

So for me, the DxO sensor score for the 5D Mark III seems to be rather close to this:

(portrait color depth bits) + (landscape dynamic range) + sqrt(sports low light ISO)
24                                                         11                                47
= 82

their score is 81 and I've rounded down (or rather truncated) two of the above numbers.

For the D800, I get 25+14+39 = 92

But using the same measure, the D3200 would only be 70 so there's obviously some compensation in their scoring for crop sensors. For the Canon G1X, I get 21+10+25=56 (not 60).

Keep in mind that I'm just making guesses here about what sort of transform they use to end up with the DxO sensor score.

1594
However on the high DR front I wonder how many get close to DR 10 and therefore would benefit from, say, a DR of 14 body over a DR of 13 body? If it isn't many then there is little point focussing on what is for most a theoretical benefit.

I for one definitely would benefit from greater DR.

Along with Lightroom 4, I think that there's a greater chance that a single photo with a DR of 14 or more would let me do some photos in one shot rather than multiple and then HDR it.

1595
At the end of the day, these are just a bunch of numbers, right?
I mean, do you think a doctor, or an engineer cares about numbers when treating a patient or building a bridge? Nah...

Numbers are important. But...knowing your serum creatinine level won't help your doctor know if your kidneys are functioning properly, without knowledge of the normal range for your age and gender.  Knowing the length a bridge must span isn't enough for the engineer to build the bridge, the load capacity must also be known. 

The DxO Score attempts to represent the performance of a sensor as a single number, a problem that's compounded when people extrapolate that number to the camera's overall performance.  To borrow Bob Atkins' apt example, that's about as meaningful as representing the Mona Lisa as its average color value:

We score many things in our life.

If you look at sports. at the Olympics, there are many sports where judges have to watch a performance and score out of 10. Do any of those scores represent the individual things that go wrong or right? No. Does anyone dispute the way in which judging is done? Not that I'm aware of.

Thus far nobody else has stepped up to measure digital camera sensors as DxO do. If DxO were doing such a bad job of it, someone else would likely come up with their own measure. Thus far I've seen nobody in this thread or anywhere else attempt to do that. To put up an argument that sensors cannot be measured is just poppycock.

As it happens, DxO publish the full test results for sensors, so that if you want to see what the components are that make up a given score, you can. It's those other numbers that are actually of interest.

1596
The point is, ther are a significant number of phtographers who regularly shoot above ISO 400, even in daylight hours. DxO scores don't take this into account properly.

What photographers spend their time shooting is irrelevant so far as DxO scores are concerned.

I aim to shoot such that I get the best image possible at all times. That necessitates keeping the ISO as low as possible.

DxO aren't reviewing cameras, they're measuring sensor performance.

Quote
Yes the D800 would benefit me as a landscape photographer, but I don't have a need to print larger than A3 most of the time

The benefit of the increased DR from the D800 is not relevant to the size of printing - megapixels is.

Quote
To be perfectly frank, impact sells more images than the dynamic range of a camera. You can use any camera from a fairly basic poimt and shoot to the most expensive large format camera to achieve impact, so the tool used becomes an irrelevance.

Right. If the subject material is poor then a camera with a better sensor won't make the photograph magically better. But on the other hand, a camera with a better sensor in the hands of someone that knows how to use it will result in a better image than they would be able to achieve with a weaker sensor.

DxO measure one aspect of the camera - the sensor performance.

When making a decision about which camera to buy, there are factors other than how the sensor performs to take into consideration. When we walk into a camera store and pick up a DSLR, we have a chance to hold it, feel it, evaluate all of the ergonomic aspects. When we're in said store, it is beyond our ability to measure sensor performance then and there. DxO empowers us by delivering to us measurements of camera sensors so that when we walk into a store to evaluate a camera to buy, we can make a decision about what's on the outside and on the inside.

1597
I think you have just elegantly put the reason why we should not be using DxO scores.

If you think that using DxO scoring that the G11 is better than the D3X is wrong - then by the same reasoning you must now understand the our scepticism over the 'D800 is better than the 5DIII' scores

I think the D800 being better than the 5DIII is evidenced in many other ways when it comes to IQ.


I am not interested in the D800/5DIII debate - rather the validity of the DxO reports.

They have clearly made you think of the possible errors by the S90/D3X comparison. If it could happen at that level then it is reasonable to question the D800/5DIII comparison.

What we can't do is pick and choose. Either DxO is misleading us or it isn't.

A lot of the activity on this forum has been based on the DxO figures and their validity.

Now we have a new set of figures saying that the entry level D3200 is as good as the pro level 5DIII.

You correctly (IMO) pointed out the stupidity of the conclusions that can be reached - perhaps this is another such case.

Ah, no. You're just not understanding what DxO are saying.

DxO isn't saying that the S90 produces better images than the D3X but that the pixels in the S90 work better than those in the D3X. To put it in car terms, the S90 gets more miles per gallon (or less litres per 100km) than the D3X because the engine is more efficient. It doesn't say that the S90 is faster than the D3X or that it can tow a larger trailer. So from a technical perspective, the S90 represents better technology at the pixel level than the D3X. Or think of it another way: if Nikon used the D3X technology to make a sensor that was the same size as that of the S90 and with the same pixel density, then that sensor would produce inferior output to that of the S90.

There is no reason why an entry level camera cannot have IQ as good as a pro-level camera except that some people have mental barriers inside their head that cannot accept it. If Nikon/Sony have a sensor that can produce better output than the 5D Mark III, are there some mystical rules that say they cannot put that sensor in an entry level sensor? No. The Pentax K5 can also produce digital images with more DR than the 5D Mark III and that's another APS-C entry level camera that is ranked alongside the 5D Mark III in terms of IQ by DxO.

1598
I think you have just elegantly put the reason why we should not be using DxO scores.

If you think that using DxO scoring that the G11 is better than the D3X is wrong - then by the same reasoning you must now understand the our scepticism over the 'D800 is better than the 5DIII' scores

I think the D800 being better than the 5DIII is evidenced in many other ways when it comes to IQ.

There have been comments from Canon owners that have bought a D800 to sit alongside their 5DIII, here at CR, about how much more usable the D800 images are, how it is hard to blow highlights and that the extra DR is amazing.

So for those that have access to both cameras, the comments that they're making about the D800's output aligns very well with what DxO have reported. Nobody that has both has said that the 5DIII has more DR, or that the images are anywhere near as usable. All of the videos that have been made by people shooting with them side by side agree with the DxO tests.

So argue away that DxO is rubbish all you like but the fact of the matter is, the D800 is producing better images than the 5DIII at ISO 100-400. Every test performed by a real person confirms this. not one person has said that the 5DIII is better than the D800 at that point in the ISO range or that the 5DIII images respond better to pulling up shadows. DxO just gives us a way to measure that difference.

1599
What is the chance of firmware update for the 5d3 and 5d2 10fps ? :)

Depending on the card you use, some people (e.g. dpreview.com) already get 24 RAW frame burst instead of Canon's stated 15 RAW frame burst. However dpreview report that the 5D Mark III is limited to 17 whilst Canon claim 18.

So on the one hand, Canon are overstating the performance (5D Mark III) which is what we'd expect, and on the other hand (7D), Canon seem to be understating the performance.

1600
Canon General / How to understand DxO scores
« on: June 08, 2012, 11:48:51 PM »
This story was published at the beginning of 2011 but it should serve to illustrate why the DxO scores are useful:

http://www.luminous-landscape.com/essays/dxomark_sensor_for_benchmarking_cameras.shtml

... as they tell an important story (once you get past the fanboyism.)

1601
I love DxO... When they announce a score, it makes all the Canon fanboys reveal themselves...

It's not your camera on top of the list? Well, then the people preparing the list are all liars & scoundrels!
At the end of the day, these are just a bunch of numbers, right?

I mean, do you think a doctor, or an engineer cares about numbers when treating a patient or building a bridge? Nah...


I love DxO when they come out with their conclusions that all things Nikon are wonderful, it makes all the Nikon fanboys come out crowing .....


Look, if DxO were such Nikon fanboys then why would they post a story saying that the S90's sensor pixels were better than Nikon's D3X?

http://www.dxomark.com/index.php/Publications/DxOMark-Insights/DxOMark-review-for-advanced-compacts

One can conclude from this analysis that the Canon G11 and S90 sensor has a much higher performance per unit of surface than the Olympus EP1 or the Nikon D3x sensors.

Now if DxO were Nikon fanboys, why would they be saying things like that?

What Canon's userbase should be saying is
"CANON - WHY THE F*** ARE YOU SCR**ING US OVER WITH YOUR DSLR SENSORS?????"

1602
If it is any solace, the D800's version of Canon's C1, C2, C3 positions on the mode dial is F_____. It's almost unusable in the way it is presented in the D800 to the point that it may as well not be there. So if you use those three positions on your mode dial, then there's nothing to consider - the D800 could have a DxOMark score of 195 and it still wouldn't make up for its crap ergonomics.

1603
I don't understand how a jump from 15 to 26 can be done with a firmware update either. Good question.


A friend and I brainstormed a few possibilities:
  • They had previously used the available RAM as a simple linear buffer, but changed their algorithm to use it as a circular buffer, allowing more images to be taken after the first was written to the card.
  • There was already a circular buffer, but the write order was optimized.
  • A redundant write verification step may have been disabled.
  • A delay that only became evident when the camera was used with higher-speed cards than were available at launch may have been discovered and eliminated.  (If you need a very fast card to get 26; otherwise, it's not this one.)
  • Someone did something really boneheaded in the original coding and it slowed everything down.


I'd vote for none of the above.

If you read the review at dpreview:
http://www.dpreview.com/reviews/canoneos7d/13
"The frame rate is impressive on itself and even more so considering that with a very fast card, such as the Sandisk Extreme Pro, in JPEG format the 7D can maintain this speed indefinitely (well, we gave up after approximately 60 sec or 320 frames) and for 24 frames when shooting RAW."

It might simply be better programming of the IO between the camera and cards.

But if the 7D can already do 24 frames with the right card now, after this new firmware is delivered, will it be higher again if you use the right kind of card?

1604
If they are not even designed to handle the same use cases, comparing the price points is meaningless and people are only doing it because they just happened to be released in about the same timeframe.

Imho "completely different usage scenarios" way of describing the 5d2 & d800 has only popped up because Canon people were desperate to save their favorite brand's newest baby's reputation. Both cameras might have different strong points and professionals will try to exploit them to gain an advantage over the competition. But for the rest of us they are certainly both complete dlsrs designed to be able to handle most situations - so any attempt to compare them seems perfectly valid to me.

5D2 and the d800 are valid comparisons as you say. Makes Nikon look like a money grabbing company to charge $1000 extra for a few mps and a little DR

D800 has 70% more megapixels than the 5D2.
D800 has 467% more focus points than the 5D2.
So a 36% price increase seems more than justified.

Only if you need them - ask the wedding togs if it is worth the extortionate Nikon price with the 100% increase in memory cards, pp time etc. We know what they will say about the poor low light performance of the D800.

Is that like how wedding photographers only need the 12MP of the Nikon D700?

When Canon are no longer selling a full frame camera with less than 30MP, will wedding photographers still be crying for fewer MP?

High mps is not the top priority I would suggest - but low light performance is for most. Like everything else enough is suficient - how many wedding togs produce huge prints? - 21mp is ideal for A3 print size.

5D2 is still THE wedding camera rather than the more expensive D700 with its inferior specs

But didn't that really cool/famous English wedding photographer ditch the 5D2 in favour of the D700 for low-light performance?

1605
If they are not even designed to handle the same use cases, comparing the price points is meaningless and people are only doing it because they just happened to be released in about the same timeframe.

Imho "completely different usage scenarios" way of describing the 5d2 & d800 has only popped up because Canon people were desperate to save their favorite brand's newest baby's reputation. Both cameras might have different strong points and professionals will try to exploit them to gain an advantage over the competition. But for the rest of us they are certainly both complete dlsrs designed to be able to handle most situations - so any attempt to compare them seems perfectly valid to me.

5D2 and the d800 are valid comparisons as you say. Makes Nikon look like a money grabbing company to charge $1000 extra for a few mps and a little DR

D800 has 70% more megapixels than the 5D2.
D800 has 467% more focus points than the 5D2.
So a 36% price increase seems more than justified.

Only if you need them - ask the wedding togs if it is worth the extortionate Nikon price with the 100% increase in memory cards, pp time etc. We know what they will say about the poor low light performance of the D800.

Is that like how wedding photographers only need the 12MP of the Nikon D700?

When Canon are no longer selling a full frame camera with less than 30MP, will wedding photographers still be crying for fewer MP?

Pages: 1 ... 105 106 [107] 108 109 ... 181