Richard Butler

Richard Butler

DPReview Administrator
Lives in United Kingdom Seattle, United Kingdom
Joined on Nov 7, 2007

Comments

Total: 5685, showing: 41 – 60
« First‹ Previous12345Next ›Last »
In reply to:

Tito Marques: Is is incorrect to compare lenses based on equivalent dof. Using the definition, focal lenght is "For an optical system in air, the distance over which initially collimated (parallel) rays are brought to a focus". That can be translated as field of view. Aperture on other hand, is "the opening of a lens's diaphragm through which light passes. It is calibrated in f/stops and is generally written as numbers such as 1.4, 2, 2.8, 4, 5.6, 8, 11 and 16. The lower f/stops give more exposure because they represent the larger apertures. This can be translated as the lens' light gathering capabilities.
Although naturally equiv depth of field as to be taken in consideration, especially depending on the type of work, thats not what the numbers on the lens primarilly mean. So, a 16mm f1.4 sigma for mft has less shallow depth of field capabilities than ff 32mm equiv 1.4 lens, but in terms of its light gathering caracteristics its still a f1.4 lens (32mm fov equiv f1.4).
Thank you.

Yes, depending on whether you use a per unit area or per image definition of exposure.

Link | Posted on Jan 12, 2018 at 19:26 UTC
On article Panasonic Lumix DC-GH5S First Impressions Review (327 comments in total)
In reply to:

mosc: I find this interresting. The sensor captures 4:3 on a roughly 2.0x crop at a resolution of 3680x2760 but the overall sensor needs to cover wider aspect ratios so the actual chip is bigger. That said, to cover DCI's 17:9 at the same ratio as 4:3, the optimal "oversized" chip is not itself a 4:3 ratio. For example using DCI's required 4096 width you get a 4:3 ratio (4096x3072) of 12.6mp but you have more height on the chip than you need for the "tallest" ratio you intend, 4:3 at a height of 2760.

The optimal ratio chip needs the width of DCI and the height of 4:3 over the same circle which is 4096x2760 and only 11.3mp. It's also a near familiar aspect ratio of 1.48x1. Right around 3:2.

If you want to make a multi-aspect ratio chip that covers 17:9 and 4:3, you should make that chip 3:2, not 4:3. If you use a 4:3 chip, it has to be 12% larger than the 3:2 would be covering the same image circle. Isn't geometry wonderful?

The 12.6MP figure is from Panasonic, I think, but I'm not convinced it's correct.

Link | Posted on Jan 12, 2018 at 19:14 UTC
On article Panasonic Lumix DC-GH5S First Impressions Review (327 comments in total)
In reply to:

Zdman: So if I look at your nice diagram of the image circle and sensor ratios I do not see how you can claim there may be a problem with focal reducers. If they were able to work with the 4:3 ratio they'll be able to work with the wider ratios as well as they all fit in the same circle.

That's a useful comment, since it prompted me to look a little bit closer at the problem, rather than just working from the numbers.

Metabones says:

"*Some EF-S and DX lenses may not have enough coverage of the entire image circle for stills after widening by 0.71x (but video will have enough coverage).*"

Which suggests the 0.71x version will usually be fine, even though it's using a fractionally-larger-than-APS-C image circle (ie it don't *always* work with the 4:3 region). So it seems my concerns should be more about the 0.71x version, which *isn't* designed to cover the full 4:3 region with an APS-C/Super35 lens.

I've updated the text accordingly. Thanks for your comment.

Link | Posted on Jan 12, 2018 at 19:04 UTC
In reply to:

anticipation_of: So Kodak is a zombie brand now. Just a name and logo that anybody with a few bucks can license and slap on their crap to help it stand out among the sea of generic, no-name garbage on Amazon. Or among the sea of me-too cryptocurrency grifters, it would seem. Sad to see the tattered remains of such an iconic organization being paraded around like this. It’s undignified.

No, there is still a company (primarily engaged in the packaging industry and manufacturing filmstock for Hollywood and Kodak Alaris) based in Rochester. However, this company (Eastman Kodak) seems to also have a fairly active licensing division.

Link | Posted on Jan 12, 2018 at 18:45 UTC
In reply to:

tedolf: Whatever happened to the concept of a Trademark being a guarantor of quality and not just source?

Is Kodak the licensor exercising quality control over its Kodak licensees?

Tedolph, Esq.

anticipation_of: I think you mean Eastman Kodak. Presumably Kodak Alaris thinks there's value to the brand name, or they wouldn't continue to licence it from EK.

Link | Posted on Jan 12, 2018 at 18:44 UTC
In reply to:

Shlomo Goldwasser: Did DPReview get a call from higher up to spin the earlier story about the mining scam?

There is no 'higher up.' Unless one of us stands on a chair.

Link | Posted on Jan 12, 2018 at 18:42 UTC
In reply to:

Tito Marques: Is is incorrect to compare lenses based on equivalent dof. Using the definition, focal lenght is "For an optical system in air, the distance over which initially collimated (parallel) rays are brought to a focus". That can be translated as field of view. Aperture on other hand, is "the opening of a lens's diaphragm through which light passes. It is calibrated in f/stops and is generally written as numbers such as 1.4, 2, 2.8, 4, 5.6, 8, 11 and 16. The lower f/stops give more exposure because they represent the larger apertures. This can be translated as the lens' light gathering capabilities.
Although naturally equiv depth of field as to be taken in consideration, especially depending on the type of work, thats not what the numbers on the lens primarilly mean. So, a 16mm f1.4 sigma for mft has less shallow depth of field capabilities than ff 32mm equiv 1.4 lens, but in terms of its light gathering caracteristics its still a f1.4 lens (32mm fov equiv f1.4).
Thank you.

The article never says the lens is anything other than F1.4.

However, in terms of behaviour (**not** just depth-of-field), it is equivalent to a 32mm F2.8 lens. Per Dr Blackjack's insightful comment below, I'm not going to turn this comment thread into re-writing out equivalence articles, but if you wish to discuss it further, you're welcome to PM me.

Link | Posted on Jan 12, 2018 at 18:34 UTC
In reply to:

Advent1sam: It would of been nice to see some examples with the g9, now the reference for m43 iq. If you get chance to make any shots with g9 would be really pleased, The em1 ii shots appeared to show the usual issues with smearing and almost missed focus too in some examples. The Sony appeared much more consistent in its output as well as the nice look to a 24mm fast wide

I'll try, as soon as Dan is finished with it.

Link | Posted on Jan 12, 2018 at 18:05 UTC
On article Panasonic Lumix DC-GH5S First Impressions Review (327 comments in total)
In reply to:

mosc: I find this interresting. The sensor captures 4:3 on a roughly 2.0x crop at a resolution of 3680x2760 but the overall sensor needs to cover wider aspect ratios so the actual chip is bigger. That said, to cover DCI's 17:9 at the same ratio as 4:3, the optimal "oversized" chip is not itself a 4:3 ratio. For example using DCI's required 4096 width you get a 4:3 ratio (4096x3072) of 12.6mp but you have more height on the chip than you need for the "tallest" ratio you intend, 4:3 at a height of 2760.

The optimal ratio chip needs the width of DCI and the height of 4:3 over the same circle which is 4096x2760 and only 11.3mp. It's also a near familiar aspect ratio of 1.48x1. Right around 3:2.

If you want to make a multi-aspect ratio chip that covers 17:9 and 4:3, you should make that chip 3:2, not 4:3. If you use a 4:3 chip, it has to be 12% larger than the 3:2 would be covering the same image circle. Isn't geometry wonderful?

If we assume this is the IMX294 sensor (which is *highly* likely), then the best details I've seen say that it's 19.2 X 13mm, which makes it nearer to 3:2 than 4:3.

Link | Posted on Jan 12, 2018 at 02:34 UTC
In reply to:

PickCellPepper: Where is the photography in this?

visualenvy - no, that's Kodak One, which is a totally separate thing. This is a different company licensing the Kodak name to mine Bitcoin. The only connection is that both companies have paid to slap the Kodak name on their product.

Link | Posted on Jan 11, 2018 at 22:25 UTC
On article Panasonic Lumix DC-GH5S First Impressions Review (327 comments in total)
In reply to:

kevin_r: Interesting how a stills camera is being morphed into a video camera. Why not simply go and buy a complete video camera to do the job? Why hang around these things hoping they'll give all the functionality of a video camera and then want to so,so occasionally take some stills? Don't talk about price since people are demanding the actual real functionality available in the video cams - but want to pay nothing for it. Go figure.

Also, anyone wanting to [use XLR with the GH5(S)](http://shop.panasonic.com/cameras-and-camcorders/camera-and-camcorder-accessories/camera-accessories/camera-external-flash-microphones/DMW-XLR1.html) can.

Link | Posted on Jan 11, 2018 at 21:44 UTC
In reply to:

jkokich: So, bitcoin says Kodak’s plan is a scam? Hmmm...

"Bitcoin" doesn't say anything. The whole idea of cryptocurrencies is that they're distributed, with no central organising body.

Then again, I'd question 'Kodak's involvement in any of this, beyond agreeing to let them stick their name on the box in return for a lisence fee.

Link | Posted on Jan 11, 2018 at 18:42 UTC
In reply to:

Thomartin: This doesn't make any sense! You rent Kodak hardware to mine Kodak-coins that I guess won't be recognized by anyone else!

This is closer to in-game virtual money that you can buy than cryptocurrency!

Go home Kodak, you're drunk.

'Kodak'

Link | Posted on Jan 11, 2018 at 18:38 UTC
On article Panasonic Lumix DC-G9 Review (922 comments in total)
In reply to:

Richard Murdey: The APSC sensor in the D500 doesn't warrant a green box in the comparison table, but having a lens-implemented VR system instead of SR gets a red flag?

I've lost faith in you guys.

You're right that that's inconsistent with our existing reviews. I've updated the table accordingly.

Link | Posted on Jan 11, 2018 at 01:50 UTC
On article Panasonic Lumix GH5S vs GH5: What's new? (89 comments in total)
In reply to:

Miki Nemeth: Excellent comparison review. It would be great to have a section about comparing the bit-rates and color depth for internal/external video. Are the two cameras exactly the same in this respect, that's why there is no a section about these?

For the most part they're pretty much the same, it's just that the GH5S gains DCI versions of the GH5's >24p UHD modes.

Going into more detail probably wouldn't fit in this slideshow, but I've put together a table [on this page](https://www.dpreview.com/reviews/panasonic-lumix-dc-gh5s-first-impressions-review/2) that shows all the GH5S's modes and highlights the ones that are distinct from the GH5.

Link | Posted on Jan 10, 2018 at 20:08 UTC
On article Panasonic Lumix GH5S vs GH5: What's new? (89 comments in total)
In reply to:

Entropy512: The "native ISO" terminology appears to come from the fact that middle grey in log footage is much farther away from where the sensor clips than it is for sRGB - and the fact that ISO ratings are defined on where middle grey lies.

For example, the A6500 has a "native ISO" of 800 when S-Log is in use - in these cases the sensor is actually running at ISO100, but the S-Log middle grey point is approximately 3 stops farther away from the sensor clip point, so Sony displays it as 800.

It appears based on the ratio of Panasonic's V-Log ISO ratings and the actual sensor ISO ratings that V-Log appears to have around one stop less headroom compared to middle grey. Of course - as soon as you're talking about log gamma curves, where the manufacturer defines "middle grey" can get to be arbitrary... The Panasonic may have one stop more shadow detail in relation to middle grey in exchange for the one stop less headroom.

It's possibly rounding or, given Panasonic told us that the number had to provide the option to shoot lower than that, it could mean (Log)ISO 320 with a tiny bit of wiggle room.

The GH5S does let you shoot VLog-L stills and Raw, so it should be possible for us to dig a little behind the scenes, once the new 14-bit Raw format is a bit better understood.

Link | Posted on Jan 10, 2018 at 18:54 UTC
On article Panasonic Lumix GH5S vs GH5: What's new? (89 comments in total)
In reply to:

Entropy512: The "native ISO" terminology appears to come from the fact that middle grey in log footage is much farther away from where the sensor clips than it is for sRGB - and the fact that ISO ratings are defined on where middle grey lies.

For example, the A6500 has a "native ISO" of 800 when S-Log is in use - in these cases the sensor is actually running at ISO100, but the S-Log middle grey point is approximately 3 stops farther away from the sensor clip point, so Sony displays it as 800.

It appears based on the ratio of Panasonic's V-Log ISO ratings and the actual sensor ISO ratings that V-Log appears to have around one stop less headroom compared to middle grey. Of course - as soon as you're talking about log gamma curves, where the manufacturer defines "middle grey" can get to be arbitrary... The Panasonic may have one stop more shadow detail in relation to middle grey in exchange for the one stop less headroom.

It's a nice theory but the minimum available ISO in VLogL mode is 320 and it's not marked as being a 'Low' setting, which implies it's based on the same hardware behaviour as ISO 160 in Standard Photo Style (this is something I hope to look into as part of the review).

This suggests 'Native ISO' isn't the ISO rating of the minimal amplification (base) setting when shot in VLog-L.

Link | Posted on Jan 10, 2018 at 18:16 UTC
On article Panasonic Lumix DC-GH5S First Impressions Review (327 comments in total)
In reply to:

liberty aint what you think: Annoyingly neither RawTherapee nor DarkTable will open raws (stills) from the GH5S.

Hope it's just a because the camera the samples came from is likely still running beta firmware.

DT manages to see the jpegs built into the raws for preview.

Or because it's the first Panasonic to put out 14-bit Raw?

Link | Posted on Jan 10, 2018 at 00:42 UTC
In reply to:

bjphotolabs: Appears like the lens used on the A7S II, in this particular test, is out performing the lens chosen for the GH5s ( in terms of overall image quality).

Any details on what lenses were used?

On the GH5S:
Panasonic 12mm F1.4
Panasonic 8-18mm F2.8-4
Panasonic 25mm F1.7
Panasonic 42.5mm F1.2

On a7S II:
Sigma 24mm F1.4 Art (EF mount)
Canon 16-35mm F2.8 III (EF mount)
Sony 50mm F1.4
Sony 85mm F1.4 GM

Link | Posted on Jan 10, 2018 at 00:41 UTC
On article Panasonic Lumix DC-GH5S First Impressions Review (327 comments in total)
In reply to:

gmke: A considerable noise has been raised over whether the GH5S sensor is derived from the IMX294CJK sensor which explicitly sports the BSI fabrication feature, and which would do more to explain deep ISO in comparison to the front side chip in the A7S-II. The full frame comparison vis-a-vis cell area alone is kind of lame in my view unless you account for the huge disparity in the percentage of cell area devoted to light gathering. Front side chips are dramatically dominated by non-sensitive circuity, so the area comparison is apples and oranges.

Having looked at the specifications, I strongly suspect the GH5S is using the IMX294 (though not the CJK variant, with its 'Quad Bayer' CFA).

However, don't forget that both of these chips have **vast** pixels in comparison with most sensors: the proportion taken up by circuitry is comparatively small, even in FSI designs.

Even when Sony first announced its move to BSI, they made clear the main performance advantages would come in small (1/1.25 and 1/1.17) sensors, which is presumably why it's taken so long for them to make their way to larger chips. The difference between the FSI 20MP 1" chip in the original RX100 and the BSI version in later models is subtle, so it's unlikely to be the apples/oranges comparison you suggest.

The sensor size comparison lets us know how much light both have access to, which then lets us get a sense for how well they're able to utilise that light...

Link | Posted on Jan 9, 2018 at 23:33 UTC
Total: 5685, showing: 41 – 60
« First‹ Previous12345Next ›Last »