dosdan

Lives in Brisbane
Joined on Dec 17, 2007

Comments

Total: 110, showing: 1 – 20
« First‹ Previous12345Next ›Last »

For years, when APS-C cameras started producing very good technical measurements and were much cheaper than FF models, FF fanbois have claimed that a big subjective difference was how the larger FF sensor reduced the quality requirements of FF lenses, so the images usually looked better.

But, now that MF cameras have become more affordable, the FF fanbois have switched from subjective image quality to the value-for-money aspect. What happened to the "larger MF sensor reducing the quality requirements of MF lenses" argument?

Strange...

Dan.

Link | Posted on Nov 14, 2017 at 23:28 UTC as 52nd comment | 28 replies

Look at the CA in the extreme left-bottom corner for the lens used with the III. Not present to the same extent in the other cameras. Check out all 4 corners.

I wonder why DPR didn't use the same lens as with the II which isn't troubled with CA?

Dan

Link | Posted on Nov 12, 2017 at 18:57 UTC as 187th comment | 2 replies
In reply to:

Sdaniella: 8mp prints = laughable "norm" !!!

why bother print at all?
one more reason "low norm" printing (opaque presentation) has zero appeal for me

if one wants to seriously print, make them at least 60" x 40" (or larger)
for gallery museum enjoyment "up close" (life size, and beyond) or eye-level street enjoyment at arms length viewing

(not wall/giant-sized billboards viewed from afar, way off beyond reach, overhead, atop skyscrapers/high hills)

i much prefer light displays, where light colors (highlight fidelity) just as important as shadow details, and i hate fake "shadows/blacks" where there is clearly plenty of light nearby

The 8MP size was chosen as the reference a fair while ago. But the reference value chosen doesn't matter that much. (Using say 16MP would just change all the normalised values the same amount.) What's important is that sensors of different MP and different physical sizes at all normalised against something. So you're comparing relative image noise performance at the same printed size, not individual pixel noise performance.

A better normalised performance usually means that an image can be printed larger.

Dan.

Link | Posted on Oct 31, 2017 at 22:52 UTC
In reply to:

Alex Georgiades: Can someone please explain to me why is there an increase in dynamic range after ISO 400? Why is there a new curve starting at ISO 500-600?

Read http://www.photonstophotos.net/Aptina/DR-Pix_WhitePaper.pdf

Sony got access to Aptina's DR-Pix tech in an IP swap.

The designer's selection of conversion gain in the pixel is a compromise between DR at low ISO and low RB at high ISO. In a "dual-gain" sensor, a pixel's CG changes at mid ISO, allowing both the best DR performance at low ISO and the best RN at high ISO.

BTW, there is a sensor design, not yet commercialised, that has triple-gain pixels.

Dan.

Link | Posted on Oct 31, 2017 at 22:40 UTC
In reply to:

dosdan: Low Read Noise/lLarge Full-Well Capacity /High DR sensors have been around for a long time. The mainstream sensor that achieved this breakthrough was the Sony IMX-071 in Aug-Sep 2010 in the Pentax K5, Nikon D7000 & Sony A-580 . Reading this 2010 article from Guillermo Luijk switched me on to the concept of an "ISOless" sensor:

http://www.guillermoluijk.com/article/perfect/index.htm

Unfortunately, photographers are fairly traditional and the concept of ISO Sensitivity on digital cameras is still strong. Nowadays, with most cameras having low RN sensors, (many are using Sony sensors), raw photographers should have an Exposure-centric mindset i.e. concentrating on capturing the greatest amount of light during the exposure, not considering the brightness of the image in LCD review screen. (The rendered brightness can be adjusted in PP.) Changing the ISO sensitivity does not increase the number of photons captured. This ultimately determines the image's SNR. It all about the light.

Dan.

As I understand it, ENOB (Effective Number of Bits) indicates the DR performance of an ADC system. A 14-bit linear raw file with a perfect sensor and ADC would have 14 stops/14-bits/14EV of engineering DR. The base DxOMark DR on the "Screen" (non-normalised) chart shows the DR at the pixel level. The ENOB of the K5 is 13.61 bits and the D7000 is 13.35 bits according to the DxoMark DR charts. This was amazing performance in 2010.

Unless the sensor itself is really noisy (not common in modern sensors), the straightness of the DR in the low ISO is an indication of low-noise ADC performance. The best sensors keep a straight line to virtually base ISO, whereas the cameras with significant ADC noise have a DR curve that flattens out at low ISO. (ADC noise is usually the most significant component in the Total RN value at low ISO. At high ISO, Sensor RN is the most significant component.)

The straighter the DR at low ISO, the closer the camera comes to being ISO-invariant .

Dan.

Link | Posted on Oct 31, 2017 at 22:19 UTC

Low Read Noise/lLarge Full-Well Capacity /High DR sensors have been around for a long time. The mainstream sensor that achieved this breakthrough was the Sony IMX-071 in Aug-Sep 2010 in the Pentax K5, Nikon D7000 & Sony A-580 . Reading this 2010 article from Guillermo Luijk switched me on to the concept of an "ISOless" sensor:

http://www.guillermoluijk.com/article/perfect/index.htm

Unfortunately, photographers are fairly traditional and the concept of ISO Sensitivity on digital cameras is still strong. Nowadays, with most cameras having low RN sensors, (many are using Sony sensors), raw photographers should have an Exposure-centric mindset i.e. concentrating on capturing the greatest amount of light during the exposure, not considering the brightness of the image in LCD review screen. (The rendered brightness can be adjusted in PP.) Changing the ISO sensitivity does not increase the number of photons captured. This ultimately determines the image's SNR. It all about the light.

Dan.

Link | Posted on Oct 31, 2017 at 22:16 UTC as 50th comment | 2 replies

I'd expect with the very small pixel size of the Q, and the limited f-number possible for the sensor's micro-lenses, that pixel vignetting would be a very serious issue with such large-aperture lenses. This is due to the large mismatch between the lens' unstopped f-number and the micro-lens' f-number.

Link | Posted on Oct 10, 2017 at 22:25 UTC as 63rd comment
In reply to:

cdembrey: Cameras do NOT make films, film-makers do! A sharp video of a fuzzy idea is still a fuzzy idea.

And at 6:01, in the frame spray from raw footage, the difference is astounding.

Dan.

Link | Posted on Sep 25, 2017 at 20:52 UTC
In reply to:

Stephen McDonald: The rolling-shutter effect that displeases me the most, is the warping or skew at the margins, with moving subjects or when panning or tilting.

It seems to be worse with cameras that have larger sensors. But with any of the Sony HX-Series cameras I've used, with "1/2.3-inch" type sensors, there's no visible rolling-shutter artifacts, when shooting in 60p.

I've read that the addition of a couple more tiny transistors on each CMOS pixel, to momentarily record the images of a global exposure, for just 1/60th-second, before each rolling scan, would produce the same effect as a global scan of a CCD.

What exactly is preventing this kind of process from being added to CMOS sensors? Is it the cost or the lack of free space to mount more transistors?

It's a loss in photo-sensitive area on the sensor's surface (a reduction in fill-factor) and also the increase in cost.

See http://www.guillermoluijk.com/article/rollingshutter/aptinaglobalshutterwhitepaper.pdf

These days, with stacked BSI sensors, you have the individual memory cells required for a global shutter implementation, one per pixel, situated on a layer behind the sensor's top surface. This means you could have a global shutter sensor without the FF reduction. But it's more expensive to do it this way.

Dan.

Link | Posted on Jul 1, 2017 at 03:02 UTC
In reply to:

stevo23: Fuji leadership always sound so sensible. Refreshing.

"Who else has lens roadmaps?"

Pentax. Here's the 645-format lens roadmap:

http://www.ricoh-imaging.co.jp/japan/products/lens/images/645_Mount_Lens.pdf

Dan.

Link | Posted on Feb 24, 2017 at 21:51 UTC
On article Prime or zoom? LensRentals investigates (237 comments in total)
In reply to:

Mike CH: Oh - subtle pokes at DxO? ;-)

DxOMark isn't a review site. The primary purpose for their camera imaging system and lens measurements are for correction of image imperfections when the raw images are developed. The publishing of their measurement data is just a bit of fluff on the side, though the publicity must do them some good.

I like individual DxOMark measurements, but I ignore their overall scoring of camera imaging systems and lenses.

However lens variation calls into question how well the measured imperfections will work with your equipment. Centering, in particular, seems problematic.

Link | Posted on Feb 14, 2017 at 21:33 UTC
On article DPReview and the TWiT Network team-up to talk cameras (24 comments in total)
In reply to:

ovatab: let's discuss if "analog" is proper term for photo-chemical process

Also, if you follow the rule of only capitalising acronyms, then wav (WAVeform Audio File Format) is a truncation, not an acronyn, and shouldn't be capitalised.

Link | Posted on Jan 30, 2017 at 22:45 UTC
On article DPReview and the TWiT Network team-up to talk cameras (24 comments in total)
In reply to:

ovatab: let's discuss if "analog" is proper term for photo-chemical process

I try to be consistent, so I always write "raw" and "URL" because the first isn't an acronym, but the second is.

Link | Posted on Jan 30, 2017 at 22:24 UTC
On article Photokina 2016: Fujifilm Interview (116 comments in total)

It will be interesting to see how Pentax responds to Fujifilm's MF if it's a successful product. Perhaps they'll try and cut down the camera size and go mirrorless and drop the price. The Pentax MF lenses are designed for a fairly deep 70.87 mm flange distance (https://en.wikipedia.org/wiki/Flange_focal_distance) which will limit how thin a MF camera Pentax can build.

Ultimately though these MF cameras are crying out for tripod use to fully realise their resolution potential, so camera bulk (or lack of it) may not be that much of a differentiator.

Competition is good.

Dan.

Link | Posted on Oct 7, 2016 at 20:07 UTC as 31st comment | 5 replies

Make sure you read this article from the guy who modified some lenses and cameras for Kubrick for this film:

http://www.visual-memory.co.uk/sk/ac/len/page1.htm

Dan.

Link | Posted on Jul 29, 2016 at 21:26 UTC as 20th comment | 2 replies
On article Medium-format mirrorless: Hasselblad unveils X1D (1179 comments in total)
In reply to:

ScanSpeak: So is this a full frame sensor?

Dr_Jon: "The Sony MF sensor is a noticeable crop from any MF film camera."

As is the sensor used in the Leica S. As is the Kodak KAF 39000 used in the Pentax 645D.

As far as I can see, in the digital era "MF" is anything bigger than FF.

Dorothy, we not in the land of film any more.

Dan.

Link | Posted on Jun 22, 2016 at 23:11 UTC
On article Medium-format mirrorless: Hasselblad unveils X1D (1179 comments in total)
In reply to:

ScanSpeak: So is this a full frame sensor?

JimW-203: "The new Hasselblad sensor is 44X33mm (so-called "crop" is actually an enlargement by a factor of 1.68). Thus, the 90mm would have an angle of view equivalent to a 54mm lens on a FF camera; the 45mm would be equal to 27mm."

Jim, you're confusing the ratio of sensor areas with CF. CF is the ratio of the FF diagonal length to the sensor's diagonal, i.e. 43.3mm/55mm = 0.78, as mentioned in the specs. The diagonal length is used so that sensors with different ARs can be compared e.g. 3:2 and 4:3.

So the AOV of a 90mm lens on this sensor will be the same as a 70mm lens on a FF camera.

This sensor has the potential to collect 0.75 stops more light at the same exposure level.

Look at the figures for the Pentax 645D which has the same sized sensor, but which has a different number of MP:

https://en.wikipedia.org/wiki/Image_sensor_format

Dan.

Link | Posted on Jun 22, 2016 at 22:21 UTC
In reply to:

endofoto: Medium formats have the lowest DR, and the worst high ISO performance. Which medium format camera are you referring to?

And I should add that the High ISO's good too with this moderm MF camera. The DxOMark Sports Score is usually (but not always) based on the highest ISO that gives an "acceptable" 30dB SNR.

The Sports Score of the 645Z is 4505 vs the D810's 2853.

Note: the "Sports Score" really doesn't indicate the suitability of a camera for sports. That's more about AF, burst speed and burst buffer depth, and the D810 is much better equipped for this role than the 645z. But with its much bigger sensor, the 645z is probably better in LL situations.

Dan.

Link | Posted on May 10, 2016 at 08:23 UTC
In reply to:

endofoto: Medium formats have the lowest DR, and the worst high ISO performance. Which medium format camera are you referring to?

Modern MF cameras with Sony CMOS sensors have good DR. The sensor in the 645Z has a 14.7 stops DR, similar to the 14.8 stops of the D810 at ISO64.

http://www.pentaxforums.com/forums/16-pentax-news-rumors/310344-pentax-645z-scores-101-dxomark.html

Dan.

Link | Posted on May 10, 2016 at 05:31 UTC
In reply to:

Prognathous: With this kind of performance, Pentax should add an option to expose-to-the-left, basically aligning exposure to the left of the histogram. The pictures may look dark before processing, but at the same time will provide the fastest possible shutter speed (for a given aperture) - without loss of detail in dark areas and without blown highlights (possibly not even light sources).

In the interests of reducing shot noise, you should always maximise the number of photons captured during an exposure. So you should ETTR, but only at base ISO.

Once the exposure results in a capture which is too dim for your liking, you normally then boost the ISO. This increases the mapping of the number of photons to a digital value representing a pixel's brightness, so the image is now brighter. But it doesn't increase the number of captured photons, so the signal-to-shot-noise ratio stays the same.

Good "ISO Invarient" performance is due to very low ADC (later-stage) noise.

What Ricoh should be doing, is offering an "ISOless" (or whatever they call it) mode for raw shooters, where the ISO either stays at base ISO, or only increases a little e.g up to ISO400 for shots that normally use high-ISO, and boosting the review image's brightness, so what's seen on the back LCD is not too dim.

This boost value would also be stored in the metadata for "default" raw conversions.

Link | Posted on May 9, 2016 at 23:43 UTC
Total: 110, showing: 1 – 20
« First‹ Previous12345Next ›Last »