pannumon

Lives in Finland Helsinki, Finland
Joined on Dec 12, 2009

Comments

Total: 227, showing: 1 – 20
« First‹ Previous12345Next ›Last »
In reply to:

OlyCzPro: Getting rid of the mirror is one thing, and this will happen whether the mount changes or not, but to make a FF or APS camera smaller whilst using the same lens mount is ludicrous.
Why not use the additional room within the camera to fill it with extras, like power, GPS and all the other bits the more compact Mirrorless cannot..

Did I read correctly though that you thought tele lenses were all the same size.. a bit of a 'stretch' when you compare the two ends of the field, Canon's 600mm f4 against Olympus' 300mm f4.. Huge difference in size..

My Panasonic 45-175mm f/4-f/5.6 is a (long) telephoto lens (equivalent to 90-350mm). Weight of 210g (0.46 lb) and length of 9cm (3.54″) means that it's totally usable even mounted on the tiniest GM1, and there IQ is great across the range. No focusing problems at the long end at f/11-equivalent aperture.

So I disagree with Richard about mixing telephotos and mirrorless cameras, although fast-aperture telephotos and 4/3 or APS-C sensor mirrorless cameras do not necessary make that much of sense.

Small and light, long reach, relatively slow-aperture telephoto lenses are great companions in good or moderate light, and things like dual I.S. on EVF can really be useful.

Link | Posted on May 24, 2018 at 13:11 UTC

So Canon actually does have an R&D department! Unfortunately, the demand for these kind of sensors is so low that they are not producing cameras using them. That 2/3" global shutter sensor seems quite interesting for action camera purposes, though.

Link | Posted on May 22, 2018 at 23:58 UTC as 30th comment | 5 replies

Sensor on the image looks quite square and smaller than in the sketch. Perhaps that is not the final sensor. A multi-aspect-ratio sensor (=oversized sensor) with additional crop mode would be very useful for video. After all, the bottleneck is the weigh of the lenses, and therefore it makes sense to use the whole image circle when shooting video. 16:9 sensor is another possibility. Very good!

I wonder if Pentax Q system would have applications for drone photography?

Link | Posted on May 16, 2018 at 18:40 UTC as 19th comment | 1 reply
On article Sony 'striping': here's the fix (791 comments in total)
In reply to:

justmeMN: The fix is to buy a different brand of camera.

@Mike CH, are you responding to me?

Considering companies not using CDAF instead of PDAF, see https://www.dpreview.com/interviews/7479700625/cp-panasonic-interview

Considering the maturity of the Sony cameras (original release), they have had at least the following issues:
-4k video that causes overheating
-on-sensor PDAF that can cause stripes
-touch screen that can only be utilized part of the time
-high megapixel count, not so good jpeg colors
-plenty of features, not so good user interface

Of course Sony is getting there, especially I believe user interface has been improved a lot (in A7/A9-series).

My purpose is not to bash Sony, but to realize that the arguably great specs in Sony cameras comes at a prize.

Link | Posted on Apr 18, 2018 at 19:10 UTC
On article Sony 'striping': here's the fix (791 comments in total)
In reply to:

justmeMN: The fix is to buy a different brand of camera.

Some brands refuse to use on-sensor PDAF, because it has issues.

+++ Sony uses cutting-edge technology +++

--- The cutting-edge technology they are using is not mature ---

The options Sony has brought us are wonderful, just don't only look at the spec-sheets.

Link | Posted on Apr 18, 2018 at 16:14 UTC
On article Video: Canon shows off its new 120MP APS-H CMOS sensor (217 comments in total)
In reply to:

pannumon: Comparing this to a 9-year old Canon Flagship model that is marketed to do 1080p does not make much sense (does 1D mk IV in practice shoot OK 720p?). And that 120Mp video does not look very sharp at pixel level either. It's probably kind of OK when downscaled to 8k, except the frame-rate makes it totally unusable.

What does this all tell us about the company?

Yeah, I wanted to be even a bit positive, but I failed.

Ahh... I finally kind of get it. They introduced this sensor in 2010 and now it actually works or is commercially available (fingers crossed), unlike in 2015/201X when they re-announced this? This kind of explains the comparison to a 2009 camera.

I have followed DPreview exclusively for the last 10 years, but there is a reason for not remembering this news from 2010. The reason is that this sensor has no real implication, except marketing.

Link | Posted on Mar 30, 2018 at 18:57 UTC
On article Video: Canon shows off its new 120MP APS-H CMOS sensor (217 comments in total)
In reply to:

pannumon: Comparing this to a 9-year old Canon Flagship model that is marketed to do 1080p does not make much sense (does 1D mk IV in practice shoot OK 720p?). And that 120Mp video does not look very sharp at pixel level either. It's probably kind of OK when downscaled to 8k, except the frame-rate makes it totally unusable.

What does this all tell us about the company?

Yeah, I wanted to be even a bit positive, but I failed.

Realtime 120Mp image processing... Maybe NSA has enough processing power to do something "useful" with that.

Link | Posted on Mar 30, 2018 at 08:21 UTC
On article Video: Canon shows off its new 120MP APS-H CMOS sensor (217 comments in total)
In reply to:

pannumon: Comparing this to a 9-year old Canon Flagship model that is marketed to do 1080p does not make much sense (does 1D mk IV in practice shoot OK 720p?). And that 120Mp video does not look very sharp at pixel level either. It's probably kind of OK when downscaled to 8k, except the frame-rate makes it totally unusable.

What does this all tell us about the company?

Yeah, I wanted to be even a bit positive, but I failed.

I think the 4k video crop factor in let's say 5D mk IV would be quite the same (although I believe the sensor area would be smaller). But honestly, I just don't get the "video" thing here. Luckily, the video is associated with Canon, not DPR.

The positive thing I was after: This looks great for stills!

Link | Posted on Mar 29, 2018 at 19:17 UTC
On article Video: Canon shows off its new 120MP APS-H CMOS sensor (217 comments in total)

Comparing this to a 9-year old Canon Flagship model that is marketed to do 1080p does not make much sense (does 1D mk IV in practice shoot OK 720p?). And that 120Mp video does not look very sharp at pixel level either. It's probably kind of OK when downscaled to 8k, except the frame-rate makes it totally unusable.

What does this all tell us about the company?

Yeah, I wanted to be even a bit positive, but I failed.

Link | Posted on Mar 29, 2018 at 18:47 UTC as 66th comment | 9 replies
On article Fujifilm X-H1 Review (1562 comments in total)
In reply to:

racin06: I'm amazed at the multitude of snide and negative comments the X-H1 is generating. Seriously, does the X-H1 ruin your day? Some of you need to get a life.

Getting offensive about fanboi comments roots from the same source as fanboism itself. I think it's fine, because emotions are related to photography and gear selection.

Link | Posted on Mar 17, 2018 at 04:36 UTC
On article Fujifilm X-H1 Review (1562 comments in total)
In reply to:

racin06: I'm amazed at the multitude of snide and negative comments the X-H1 is generating. Seriously, does the X-H1 ruin your day? Some of you need to get a life.

Personally I just enjoy discussing about camera trends (or cameras I can't afford).

I don't see many people moving to Fuji because of X-H1. It's not enough to be as good as others, it's often not even enough to better (Samsung NX1). Fuji has great AF and video, but it takes time to gain market trust. If Fuji is persistent enough to continues improving their cameras as they have been so far, I believe X-H2 or X-H3 will hit the jackpot.

Link | Posted on Mar 16, 2018 at 16:25 UTC

In case someone else is wondering: "Canon EF, Nikon F, Pentax K & Sony FE mounts are available" (their website). Furthermore, as mentioned in the article, Micro Four Thirds and Fuji X mount adapters are available for a very reasonable $30 price.

If the lens is optically excellent, mounting it to APS-C or 4:3 sensor camera might even make sense.

Link | Posted on Mar 7, 2018 at 21:52 UTC as 24th comment
In reply to:

pannumon: 10 000 calculations / second is not much, unless the formula is complex. 10 000 simple calculations / second = 0.01 millions of instructions per second (MIPS).

Reference 1: Intel 8080 from 1974 could already do 0.4 MIPS. It was not the fastest processor at the time, but cheap enough that it was used in home micro computers.

Reference 2: iPhone 6 from 2014 can do over 3 000 MIPS, and that's yesterday's technology already.

I can't see how zebra warning would use much processing power. It's extremely simple task to do. In the simplest implementation it is not even necessary to check all the pixels, just the pixels that are shown in the screen. The best of course would be to check each color channel for each pixel.

@ovengloves, I agree that I am simplifying. The idea of the calculations was just to put things on scale, and more specific, to get a rough idea what is possible and what is not. The ratio between 10 000 (calculations per second) and 1 000 000 000 (instruction per second) is the same as between 1 and 100 000! The difference is not two-fold, not 10-fold, not 1000-fold, but 100 000-fold!

First I thought that the lack of processing power CANNOT be the problem with zebra warning, but based on the simple calculation, I now acknowledge that it CAN be a problem.

Link | Posted on Feb 22, 2018 at 08:33 UTC
In reply to:

pannumon: 10 000 calculations / second is not much, unless the formula is complex. 10 000 simple calculations / second = 0.01 millions of instructions per second (MIPS).

Reference 1: Intel 8080 from 1974 could already do 0.4 MIPS. It was not the fastest processor at the time, but cheap enough that it was used in home micro computers.

Reference 2: iPhone 6 from 2014 can do over 3 000 MIPS, and that's yesterday's technology already.

I can't see how zebra warning would use much processing power. It's extremely simple task to do. In the simplest implementation it is not even necessary to check all the pixels, just the pixels that are shown in the screen. The best of course would be to check each color channel for each pixel.

The point was simply to note that 10 000 calculations per second does not actually mean anything, just as CaPi noted, and that regarding to computing power, the number itself is low. A single calculation needs to have tens of thousands of instructions in order to have effect on camera performance (assuming that the CPU power is distributed somewhat efficiently between the tasks). If the CPU load would constantly be high due to IBIS, that would mean fast battery drainage, but I don't think that is the case.

Actually, opposite to what I wrote earlier, checking 24 million pixels 60 times per second can actually take a lot of computing power (24M x 60 = 1.44 G calculations/second). However, in practice it is not necessary to check every pixel 60 times per second.

Link | Posted on Feb 21, 2018 at 19:42 UTC

10 000 calculations / second is not much, unless the formula is complex. 10 000 simple calculations / second = 0.01 millions of instructions per second (MIPS).

Reference 1: Intel 8080 from 1974 could already do 0.4 MIPS. It was not the fastest processor at the time, but cheap enough that it was used in home micro computers.

Reference 2: iPhone 6 from 2014 can do over 3 000 MIPS, and that's yesterday's technology already.

I can't see how zebra warning would use much processing power. It's extremely simple task to do. In the simplest implementation it is not even necessary to check all the pixels, just the pixels that are shown in the screen. The best of course would be to check each color channel for each pixel.

Link | Posted on Feb 21, 2018 at 15:40 UTC as 21st comment | 6 replies
In reply to:

ewelch: Panasonic announced a similar chip before Sony. So Sony is not the first.

And the Panasonic sensor is 8K.

@sibuzaru Are you seriously saying that 1/3 stops quantum efficiency is more important than 36Mp vs 1Mp.

Link | Posted on Feb 17, 2018 at 00:22 UTC
In reply to:

Sergey Borachev: Pretty good as a cheap mid-range camera,
EXCEPT for this deal breaker:

- 16:9 sequential field EVF!!!

If you are still reading, then realise that it's old underperforming parts that are used, e.g. 4 stop only "Dual-IS", 1/4000 max speed, most probably the old DFD, a kit lens with very ordinary performance, no 60p video, but it's only $999.

Not an exciting camera at all, but a pretty good camera if you are not demanding in your requirements but have to have a tilting EVF and if your eyesight is good. Just as previously feared, Panasonic is again ruining its camera with lousy EVFs.

If you have to buy something cheap and small, Fuji's 24MP A5 is much better.

X-A5 should be compared against GX80/85, not GX9.

Your main criticism for GX9 is the quality of EVF. In the next sentence you instead recommend a camera that does not have an EVF in the first place. GX85 also has EVF.

Another criticism is the 4k video frame rate, which is much better than in the camera you recommend. Also GX85 has the same video specs as GX9.

Then you criticize the Dual I.S. being older version. Which version does the Fuji X-A5 have? GX85 supports dual I.S.

Then you say that the kit lens has "very ordinary performance", although in fact it is a very good kit lens, with versatile 24-120mm-equivalent range.

Max mechanical shutter speed of 1/4000 is the same for GX9 and X-A5 (and GX85).

Fuji A-X5 probably beats the "Old DFD" in AF-C, whereas any DFD will beat A-X5 in AF-S.

Fuji has slightly larger sensor and lenses, and more megapixels.

That being said, I'm quite disappointed about GX9. It is not so different from GX85, that has much better value for money.

Link | Posted on Feb 13, 2018 at 23:41 UTC
In reply to:

Mark300: Panasonic - shutter problem

Guys be very careful while you purchase pana camera cause some of them is affected with shutter problem. Mechanical shutter works up to 1/1000 sec not faster. When you set lets say 1/4000 camera take it into consideration but mech. shutter still work with 1/1000 - slower than expected = shots are overexposed.

See detail and compare difference between electronic (not afected) and mechanical (afected)

https://www.dpreview.com/forums/thread/3962043

Rgds.

Some GX8's had this issue in 2016. All the new Panasonics have different shutter mechanism. All major camera manufacturers have quality control issues, and Panasonic seems to be one of the most reliable ones.

Link | Posted on Feb 13, 2018 at 23:15 UTC
In reply to:

Sergey Borachev: Pretty good as a cheap mid-range camera,
EXCEPT for this deal breaker:

- 16:9 sequential field EVF!!!

If you are still reading, then realise that it's old underperforming parts that are used, e.g. 4 stop only "Dual-IS", 1/4000 max speed, most probably the old DFD, a kit lens with very ordinary performance, no 60p video, but it's only $999.

Not an exciting camera at all, but a pretty good camera if you are not demanding in your requirements but have to have a tilting EVF and if your eyesight is good. Just as previously feared, Panasonic is again ruining its camera with lousy EVFs.

If you have to buy something cheap and small, Fuji's 24MP A5 is much better.

I mostly agree, but his camera does shoot 60fps 1080p video, and of course 24/30fps 4k (Fuji X-A5 does not shot 60p 4k video either, in fact the 15fps 4k of Fuji X-A5 should not be considered 4k at all).

The 12-60mm kit lens is actually very good (for a kit lens).

After the initial disappointment, if the price soon drops to $150-$200, I think G9 will probably have quite good value-for-money.

Link | Posted on Feb 13, 2018 at 20:42 UTC
In reply to:

chaeric2: I don't know what the terminology is for rally drivers, but for motorcycle track days we would call where that photographer was standing the "crash zone."

Also, I find it interesting that spectators are so visible and close by to the path the rally driver is taking. Photographers at motorcycle events were instructed to make an attempt at camouflage and to not attract attention so the motorcycle rider wouldn't be distracted at speed.

@Yuktuk. That video is great. Especially the part from 02:20 to 03:30 is unimaginable, and relates to the topic very well.

Link | Posted on Feb 3, 2018 at 12:51 UTC
Total: 227, showing: 1 – 20
« First‹ Previous12345Next ›Last »