Henrik Herranen

Lives in Finland Tampere, Finland
Works as a Digital Signal Processing Software Engineer, MSc
Joined on Oct 6, 2005
About me:

Plan: To baldly shoot what everyone has shot before.

Comments

Total: 397, showing: 1 – 20
« First‹ Previous12345Next ›Last »
In reply to:

D logH: Well, with a 72mm filter thread, I would expect mechanical vignetting to be quite obvious on a 35mm frame wide open. I am curious where the 79mm entrance/exit pupil would be to give this an f-number of 0.95? Perhaps DPreview could contact the company to explain this odd discrepancy.

All right, then everything seems to be in order - except that DPreview copied and pasted the impossible numbers without asking themselves "how come?"

Link | Posted on Apr 27, 2018 at 17:28 UTC
In reply to:

D logH: Well, with a 72mm filter thread, I would expect mechanical vignetting to be quite obvious on a 35mm frame wide open. I am curious where the 79mm entrance/exit pupil would be to give this an f-number of 0.95? Perhaps DPreview could contact the company to explain this odd discrepancy.

Well, with a 72 mm filter thread it is physically impossible for the lens to be a 75mm f/0.95, because the smallest theoretical front element diameter would be 75mm/0.95 = 79mm. And, as can be seen from the front photo, the front element is significantly smaller than the filter thread, i'd guess about 15%. So, one has to assume this lens has a black hole or some other game-changing technology inside it.

Damien Demolder: Should your article perhaps comment on this apparent discrepancy of numbers?

Link | Posted on Apr 27, 2018 at 06:11 UTC

Lars Rehm: I think "640 x 380 pixel resolution" should read "640 x 480 pixel resolution".

Link | Posted on Apr 26, 2018 at 04:18 UTC as 33rd comment | 1 reply
On article Quick look: Canon's new compressed Raw format (241 comments in total)
In reply to:

John C Tharp: I've been curious as to where Canon was going with CR3- thanks Jeff and Carey for digging in!

Except that the premise of the article is flawed. They start by claiming CR2 is uncompressed, which just isn't true.

Link | Posted on Apr 2, 2018 at 21:06 UTC
On article Quick look: Canon's new compressed Raw format (241 comments in total)
In reply to:

girlperson1: Compressed RAW..... The ultimate oxymoron of 2018!!

"Compressed RAW..... The ultimate oxymoron of 2018!!"

This is incorrect - just as the whole premise of the article, BTW.

Canon has used losslessly compressed RAW forever, or at least since my first DSLR, the 5D in 2005. Lossless compression is beautiful in that it takes advantage of redundancy in the original information so that it can (almost always) make files smaller, but still reconstruct the original information bit-by-bit, hence the word lossless. Just like if you zip a text file: you lose 50-80% of the file size, yet you can restore every character of text without losing a single bit of data.

Link | Posted on Apr 2, 2018 at 21:01 UTC
On article Quick look: Canon's new compressed Raw format (241 comments in total)
In reply to:

entoman: The most important thing to note here is that you have a choice. The M50 (and presumably all future Canon cameras) gives you the choice between enabling compressed CR3, or default uncompressed CR2 files.

I would imagine that after a few iterations, the CR3 option, if it proves popular, will become the default setting, but that we'll still be able to opt for CR2 if we wish. Also it's almost guaranteed that if you have a (future) Canon with twin card slots, that you'll be able to save CR2 to one card, and CR3 (or JPEG) to the other.

CR3 will save storage space, speed up write-to-card times and enable more prolonged bursts as the buffer will take longer to fill. It will probably also enable faster burst speeds than CR2.

Another major advantage is that during post processing, files will render and import/export faster, improving workflow.

There is no such thing as "uncompressed CR2 files". Have a look at two CR2 files: are they the same size or of different sizes? Yes, right, they are differently sized. How could that be if they weren't compressed?

This article is broken to begin with.

Link | Posted on Apr 2, 2018 at 20:51 UTC
On article Quick look: Canon's new compressed Raw format (241 comments in total)
In reply to:

Alex Tutubalin: Link to zip with RAWs is broken :(

Carey: The premise of the article is also broken. As opposed to what this article claims at its very beginning, CR2 files are definitely compressed. That's one of the beautiful things of the (CRW and) CR2 format(s): because they are compressed losslessly way, they are smaller than dumb uncompressed raw files.

Link | Posted on Apr 2, 2018 at 20:49 UTC
On article Quick look: Canon's new compressed Raw format (241 comments in total)
In reply to:

shadowz: A BIG THANKS to DPR for bringing this issue to the table for analysis !

I had wished that this be discussed and had even spelt out my thoughts at the comments section of M50 sample gallery .
I am delighted that the topic is generating so much interest & discussion ....
regards

Except that the whole article is WRONG to begin with. CR2 files, just as CRW files before them, have always been compressed. Claiming CR3 files are the first ones to be compressed is simply and plainly WRONG. And with that, the whole article is irrecoverably (sp?) flawed, misleading, incorrect, wrong, wrong, wrong.

Link | Posted on Apr 2, 2018 at 20:46 UTC
On article Quick look: Canon's new compressed Raw format (241 comments in total)
In reply to:

derfotograf: First they crank up the resolution (MegaPixels!) and charged for every upgrade.
Now they degrade the image quality with useless worm patterns and try to charge the consumer again.

How weird is the world... time to boycott the camera manufacturers!

Barney: They claim that CR2 isn't compressed right at the beginning of the article. This is simple and plainly WRONG. CR2 files ARE compressed, why else would each and every file be of different size? Please, see to it that this deeply flawed article is retracted!

Link | Posted on Apr 2, 2018 at 20:44 UTC
On article Quick look: Canon's new compressed Raw format (241 comments in total)

"Older Canon's that used the CR2 file format could capture either uncompressed Raw files or 'medium' or 'small' equivalents, which saved you disk space at the expense of reducing resolution."

Wrong, wrong, WRONG! And in a much worse way that the plural of Canon, "Canons", is written without an apostrophe!

CR2 files are definitely compressed. In a lossless way, like a zip file, where you can retract every bit of original information. If CR2 files weren't compressed, why would file size depend on the contents/complexity of the scene? The whole premise of this article is deeply flawed.

Carey Rose, Jeff Keller, I respect you, but this time you have made a REALLY huge mistake. Please retract this article, as it is so fundamentally incorrect to begin with!

Link | Posted on Apr 2, 2018 at 20:42 UTC as 35th comment | 3 replies
In reply to:

Rishi Sanyal: Just a quick note to those wondering if you're better off shooting ISO 640 over ISOs between 200-500. This is a good question.

The answer is **No: not if you have enough light to expose ISOs 200-500 properly**.

Recall that dynamic range is not everything, and generally the more light you collect, the better your image, because the higher the signal:noise ratio (because of [photon shot noise](http://bit.ly/shotnoise)).

If you can set a shutter speed/aperture to expose, say, ISO 320 properly, you should *not* rather choose ISO 640 and shorten your exposure (to preserve highlights that the higher amplification of ISO 640 might clip). That would mean lower overall signal:noise ratio, and would essentially have the same overall effect of shooting with a smaller (in this case, roughly APS-C) sized sensor.

Dual-gain helps low light performance, & shouldn't affect your exposure decisions any differently. Other than perhaps biasing toward ISO 640 rather than 500 in low light.

Rishi: "Recall that dynamic range is not everything, and generally the more light you collect, the better your image, because the higher the signal:noise ratio (because of photon shot noise)."

While this is generally true, it is not true anymore if a sensor has such a gain structure that it can give higher dynamic range at a higher ISO value. That's where the "generally" breaks down.

But, as I said in my previous message above, don't take my word for it. Please test it yourself and report your findings. I believe we'd all like to see what actually happens!

Link | Posted on Mar 17, 2018 at 22:11 UTC
In reply to:

Rishi Sanyal: Just a quick note to those wondering if you're better off shooting ISO 640 over ISOs between 200-500. This is a good question.

The answer is **No: not if you have enough light to expose ISOs 200-500 properly**.

Recall that dynamic range is not everything, and generally the more light you collect, the better your image, because the higher the signal:noise ratio (because of [photon shot noise](http://bit.ly/shotnoise)).

If you can set a shutter speed/aperture to expose, say, ISO 320 properly, you should *not* rather choose ISO 640 and shorten your exposure (to preserve highlights that the higher amplification of ISO 640 might clip). That would mean lower overall signal:noise ratio, and would essentially have the same overall effect of shooting with a smaller (in this case, roughly APS-C) sized sensor.

Dual-gain helps low light performance, & shouldn't affect your exposure decisions any differently. Other than perhaps biasing toward ISO 640 rather than 500 in low light.

Rishi: I think you might be wrong here.

If ISO 640 has more dynamic range than ISO 320 when image brightness is matched (you use half the exposure to match for higher ISO), and this hasnt been achieved with trickery like spatial noise reduction, then it necessarily and automatically follows that ISO 640 has better SNR = less noise than ISO 320. It might be against your common sense, but so it will be.

You don't have to take my word for it. Actually, I ask you not to. I (and I believe many others) would much appreciate a test where you'd shoot the same scene with ISO 320 at f/x and t=1/y, then with ISO 640 at f/x and t=1/2y. I honestly believe that the result would surprise you. As a professional in signal processing, I am quite sure that the ISO 640 result will look better, with less noise both in highlights and shadows. (Assuming your basic assumption is correct, i.e. that dynamic range at ISO 640 is greater than at ISO 320.)

Link | Posted on Mar 17, 2018 at 22:07 UTC
In reply to:

tedolf: 25mm is awfully short for a macro lens.

What is the working distance from the front element?

tEdolph

How about searching for the words "working distance" in the article?
"[...] which allow for a max working distance of 40 mm (5x) to 45 mm (2.5x)."

Link | Posted on Mar 7, 2018 at 20:46 UTC
In reply to:

Sunshine7913: Can it shoot full body pic of huge beetle?

No it can't, because minimum magnification is 2.5x. Any in-focus bug must be less than 14 mm long/wide and less than 10 mm tall if you use a full frame camera. And naturally even less if you use a smaller sensor. To shoot larger subjects, use a normal macro lens that can focus to infinity.

Link | Posted on Mar 7, 2018 at 20:43 UTC
In reply to:

Snoddas: “CFExpress is essentially the next revision of XQD, and there should be full backward compatibility with XQD, and that getting D4/D5/500/D850’s to work with CFE cards should be a simple software patch.”

That quote defeats itself. Where I come from, "full backward compatibility" means that the product should work with old hardware as-is. If it needs a software patch, it by definition isn't full backward compatible.

Link | Posted on Mar 6, 2018 at 22:02 UTC
On article Canon EOS M50 Review (1296 comments in total)

Dan Bracaglia: You write: "The M50 is the first Canon to use the new CR3 Raw format, which has an updated compression option called C-Raw (compressed full resolution, rather than the downsized 'Small' and 'Medium' Raw formats)."

CR2 is already always compressed, and it also offers full resolution, so I'm not really sure what makes C-Raw different. Could it be that C-Raw is compressed in a lossy way (CR2 files are always compressed losslessly)? Or is the difference somewhere else? What am I missing here?

Edit: Found it in another piece of DPreview news: "The M50 is also the first camera to offer a new CR3 Raw format. It provides a standard "image quality priority" Raw setting as well as a "size priority" C-Raw that produces up to 50% smaller files with the same resolution, albeit with a slight drop in image quality."

So, seems like lossy compression. Might be a good option, as long as the format is reverse engineered by the open software developers as soon as possible.

Link | Posted on Feb 26, 2018 at 09:23 UTC as 298th comment | 1 reply
On article Yongnuo announces YN 14mm F2.8 in Canon mount (153 comments in total)
In reply to:

CaPi: Amazing that Canon customers seem to need even more lens vendors. ;)
I was under the impression that they already had the widest variety to choose from.
Btw I would think that the featured offering would compete with other 3rd party lenses rather than oem equipment

CaPi: Being the market leader has its perks: a small piece of a big market is more than a huge piece of a niche market.

Link | Posted on Dec 30, 2017 at 11:50 UTC
On article How to photograph the northern lights (80 comments in total)
In reply to:

John Koch: The camera gear and settings are feasible. To attain the necessary latitude settings, however, can be expensive and chilly. The nights may be long, but cloudy. The short days in between shots must be very dim and bleak. Unless you live near the poles, air fare to Iceland may make it the most affordable option. Odd that no UK/Scot locations are listed.

Will B Milner and Barleyman: And the Scots would be wrong - at least most of the time.

The Northernmost corner of mainland Scotland is below the 59th latitude, which is pretty low. Heck, even Helsinki is to the North of that, and Helsinki is at the southern end of Finland!

I live at the 62nd latitude (200 km to the North of Helsinki). From time to time there are nice Northern Lights here - so good in fact that I took some completely acceptable pictures with my Canon PowerShot G2 in 2002. Two years ago my wife took some with her Samsung phone, and I managed to get a few with my 5D2 from an aeroplane below the Helsinki latitude!

Nevertheless, chances for nailing nice and bright, full-sky Northern Lights get /hugely/ better if I drive 800-1000 km North to get above the Arctic Circle. There, you get on average up to 200 nights per year of Auroras.

Link | Posted on Dec 15, 2017 at 19:23 UTC
In reply to:

Greg VdB: All fine with me, as long as Roger Cicala keeps having enough time to produce his fabulous test reports.

Roger, I think you should make "QA and testing Czar" your official title!

Link | Posted on Dec 7, 2017 at 07:26 UTC
In reply to:

semorg: a nice and informative writeup!
There is a lot of unit variations with Samyang 14/2.8. I ordered like 4 and one was perfect. That copy is very sharp edge to edge. There is vignetting at 2.8. But I tested it on my book case and it has amazing resolving power to extreme edge. What it lacks is nice coating and a strange color tint.

Why would you want a strange color tint?

Link | Posted on Dec 2, 2017 at 21:35 UTC
Total: 397, showing: 1 – 20
« First‹ Previous12345Next ›Last »