D850: The voice of Brad Hill

[Fanzizou]

Le niveau est "top of the hill" ici 

[Fogger]

Il prend chère le pauvre.

[jeanbart]

Je ne connaissais pas ce blog, en tout cas il y a des piafs super colorés dans son coin. 

[Verso92]

Il prend chère le pauvre.

Que veux-tu dire ?

[Fogger]

Que veux-tu dire ?

C'est en rapport avec la première page du fil, si le pauvre Brad est passé par ici, il a du se poser des questions...
J'espère qu'il a de l'humour.

[4mpx]

Je ne connaissais pas ce blog, en tout cas il y a des piafs super colorés dans son coin. 

Je connais son site depuis dix ans. J'aime beaucoup ce qu'il fait.

[Verso92]

C'est en rapport avec la première page du fil, si le pauvre Brad est passé par ici, il a du se poser des questions...
J'espère qu'il a de l'humour.

Tu voulais donc dire qu'il prend cher...

(je pensais qu'il y avait une allusion cachée dans l'orthographe erronée...)

[Fogger]

Ah oui, grosse faute effectivement.

[Verso92]

Ah oui, grosse faute effectivement.

Oui.

[Fogger]

Je connais son site depuis dix ans. J'aime beaucoup ce qu'il fait.

Il y a presque dix ans, il a du prendre son pied ici, au D3.
http://www.naturalart.ca/galleries/birds/inflight/detail_14.html

[jeanbart]

Je connais son site depuis dix ans. J'aime beaucoup ce qu'il fait.

Son site est vraiment très bien fait, j'aime beaucoup les onglets qui permettent d'associer du commentaire aux images.

[big jim]

Premiers retours de Brad de son tour dans la Great Bear Rainforest... Je vous laisse les découvrir sur son blog 

[4mpx]

Premiers retours de Brad de son tour dans la Great Bear Rainforest... Je vous laisse les découvrir sur son blog 

Je savais déjà ce qu'il allait dire. 

[big jim]

Je savais déjà ce qu'il allait dire. 

En tous cas, c'est très bon signe, et même si je lis les tests avec attention, les retours "terrain" de personnes pointues apportent autre chose que les courbes 
Et même s'il faut prendre un peu de recul parfois, de ton côté, tu arriverais à tirer des images sharp avec un cul de bouteille asthmatique + TC 

[Fogger]

En tous cas, c'est très bon signe, et même si je lis les tests avec attention, les retours "terrain" de personnes pointues apportent autre chose que les courbes 
Et même s'il faut prendre un peu de recul parfois, de ton côté, tu arriverais à tirer des images sharp avec un cul de bouteille asthmatique + TC 

C'est le genre de personne qui devrait être invité sur les plateaux de présentation.
Trois photographes de mode pour la présentation c'est bien mais un peu de trop.

[77mm]

"For now all I'll say about the D850 is that it DEFINITELY exceeded my expectations, both in terms of ISO performance and overall image quality. While the BEST wildlife camera offered by Nikon for my uses is still definitely the D5, I wouldn't hesitate to describe the D850 as the most versatile DSLR developed to date (by any manufacturer) for the serious nature photographer."

http://www.naturalart.ca/voice/blog.html

Je m'attendais à ces conclusions compte tenu de l'environnement dans lequel il travaille mais c'est toujours intéressant de le lire. 

[Fogger]

La suite
14 Octobre
http://www.naturalart.ca/voice/blog.html

[ORION]

I'm "rushing" this blog entry out after shooting with the Nikon D850 for only 3 days and primarily to stem the email avalanche I'm getting about the camera! During these 3 days I've shot a little over 2,000 raw images. I've done a mix of methodical testing (mostly ISO performance testing) and "just shooting" in scenarios that put several aspects of the camera's capabilities (such as AF performance) to informal tests. You know, things like shooting a running Portuguese Water Dog, et cetera! Most everything I touch on below will be discussed in much more detail (and often with example images) in future blog entries. My NEXT blog entry will go into WAY MORE detail on one aspect of ISO performance, specifically documenting how visible noise changes with increasing ISO (and it will include comparisons to the D500, D800e, and D5).

Note that my D850 arrived without the MB-D18 Battery Grip (they aren't shipping in Canada yet) so SOME of my D850 testing will be delayed until the grip arrives. This includes things like examining the lowest shutter speeds that long telephotos can be hand-held at (compared to other lower-resolution cameras such as the D5) where use of the battery grips makes for more of an "apples-to-apples" comparison.

So...with no further ado:

OVERALL ONE-SENTENCE SUMMARY: The Nikon D850 is an EXCEPTIONAL high-resolution, low-ISO DSLR.

Some specifics:

1. Exceptional tonal range and "drop dead" fantastic overall image quality at low ISO's (ISO 64 to about ISO 250). After just over 2,000 shots over a range of conditions (and ISO values) there is no doubt in my mind that this is a superb landscape camera that exceeds the image quality of the D800e and/or D810 by a significant margin. I haven't tested (or have any way to test) dynamic range directly, but it seems exceptional at low ISO's (it seems almost impossible to blow out a highlight!). But note that this is a fully SUBJECTIVE statement based only on my experience (and years of looking critically at images). Based on what I am seeing I suspect low ISO dynamic range will prove to be very high on the D850 (as will tonal range and colour depth at low ISO's).

2. Autofocus: Exactly as advertised - D5 quality (so a step above D500 quality in almost all ways EXCEPT in viewfinder coverage). I am a huge fan of the 9-point Dynamic Area mode that is found on the D5 (but NOT on the D500) and it works just as well on the D850 as it does on the D5. Exceptionally high "hit rates" on moving subjects (similar to what you'd get with the D5). At this point the more demanding sensor of the D850 (compared to the D5) does not seem to be causing the AF system to even hiccup (this was a concern I had but had not previously discussed online).

3. Mirror blackout - and between-frame image stability (within bursts). These are both functions of the mirror driving system and the D5 beats the D500 noticeably here (and both the D5 and D500 had "new" mirror-drive systems when they were introduced). The D850? Seems as good - or very close to as good - as the D5 (I think it's possible they used the same new mirror driving system as the D5). Not many folks comment on this feature, but it does make a pretty big real-world difference when shooting action. Despite the slower frame rate, the D850 "feels" like the D5 (and better than the D500) when shooting action.

4. Burst Rate: NOT exactly as advertised. All Nikon literature has really said is: "Despite the heavy load, the camera is capable of continuous shooting for up to 51 frames (body alone) even in a 14-bit lossless compressed RAW" (presumably at fastest frame rate?). What have I found when shooting full-size 14-bit uncompressed raw files? When using a Lexar Professional XQD 2933x card (440 MB/s) I get approx 40 shots at 7 fps before the camera slows to about 3 fps. But it then chugs along at that rate almost indefinitely. BUT...when using a Lexar Professional 2000x SD card (UHS-II and 300 MB/s) I get about 24 shots at 7 fps before slowdown to 3 fps (and then it chugs along for as many frames as you want). So...I'm getting MORE than 51 raw images per burst, but NOT at the maximum frame rate during the entire burst (even when using the fastest XQD card currently available). Hmmm...

What about 20 frame bursts - how many of those (with about a 1 second gap between them) can you do? This tends to be how I (and I think a lot of photographers) shoot action - repeated bursts separated by a second or two. Here's what I'm getting:

A. With XQD card (same card as described above): I get two 20-frame bursts at 7 fps and THEN it slows down (to about 3 fps) just a few frames into the 3rd burst.

B. With fast UHS-II SD card (same card as described above): Just ONE 20-frame burst  [at]  7 fps and then it slows right down (to about 3 fps) a few frames into 2nd burst.

So...burst rate is probably adequate for most uses and most users...but some sports shooters and bird-in-flight types might not be too happy with it. And...those concerned about burst depth and shooting repeated bursts should get and use a fast XQD card.

5. ISO performance? BELOW my expectations (which were conservative).

IMPORTANT NOTE: Expanding on this topic - including giving comparative samples of the images upon which the following statements are made - is the entire focus of my next D850 blog entry. In that entry I will disclose all the gory details of my testing. And note that these comments are based upon viewing RAW images, not in-camera JPEG's.

OK...In judging ONLY visible noise (not dynamic range or colour depth) in raw files, the D850 does not match the D800e, D500, or (of course) the D5 in noise characteristics at moderate to high ISO's. Note that at this point I have tested the ISO on all 4 cameras on two very different scene types but got identical results.

How did the D850 actually stack up against the other cameras? Consider a D850 raw image file captured at ISO 3200. At what ISO in the "other" cameras do you see comparable noise?

A. With D500? ISO 3200 to ISO 4000 (so just very slightly better than the D850 - in the 1/3 stop range at most).
B. With D800e? D800e: ISO 4000 and in some shots ISO 5000 (so noticeably better than the D850). Note that the D800e and D810 are virtually identical in this regard.
C. With D5? ISO 8000 to 10,000 (in a class of its own...which isn't at all surprising when you compare the pixel pitch of the cameras being tested).

Another observation was very obvious when I was going cross-eyed looking at all these images - the D850 seems particularly high in luminosity noise.

Another way of looking at ISO performance is examining the ISO where one first can see colour noise or luminosity noise (when viewing raw files at 100% magnification on a 110 ppi monitor). So...

A. When does colour noise first become visible on the various cameras (i.e., at what ISO do you need to begin suppressing colour noise during raw processing if you care about producing noise-free full-resolution images when viewed at 100% magnification)?

• With D850: At ISO 400
• With D500: At ISO 640
• With D800e: At ISO 400
• With D5: At ISO 1000

B. When does luminosity nose first become visible and require suppression?

• With D850: At ISO 400
• With D500: At ISO 640
• With D800e: At ISO 640-800
• With D5: At ISO 1600

Note that SUBJECTIVELY I have noticed already that ISO 3200 shots taken with the D850 when "just shooting" appear quite flat (i.e., with a narrow tonal range) - in a sense quite similar to D500 ISO 3200 shots (and not nearly as appealing tonally as 3200 ISO images taken with a D5).

For me (as a wildlife photographer) the biggest take-home lesson is that the D850 and D500 are exceptionally close in the amount of noise they exhibit at various ISO's (with the D500 having only a VERY small edge over the D850...in the "1/3 stop at most" range). While this may disappoint some (and doesn't come close to matching some of the early and unrealistic marketing "hype" about the D850), I personally think that in an absolute sense the D850 has outstanding ISO performance for a 45.7 MP DSLR.

6. And two other "little" things: Incredibly bright viewfinder (best I've ever seen). AND...battery life (using EN-EL15A) seems great - got over 1800 shots on first battery and was doing a lot of menu stuff, image review, and Live View shooting. You should be able to shoot forever with the EN-EL18b battery in the battery grip!

Stay tuned...my next D850 blog entry will get into the nitty-gritty! And you'll start reading comments about what the ISO performance of the D850 really MEANS in terms of how suitable of a camera the D850 is for wildlife photography (HINT: Don't expect me to proclaim the D850 as the "Best Wildlife Camera Ever!"). ;-)

Cheers...

Brad

Feedback to: feedback [at] naturalart.ca

Link directly to this blog post: http://www.naturalart.ca/voice/blog.html#NikonD850_FirstImpressions

Tu peux me le traduire en breton STP.
Merci

[suliaçais]

ou en GALLO car on est en pays GALLO par cheu moi......

[JMS]

...en français selon le traducteur Google...

"RÉSUMÉ GÉNÉRAL D'UNE PHRASE: Le Nikon D850 est un DSLR EXCEPTIONNEL à haute résolution et à faible ISO."

Ensuite il y a pas mal de commentaires sur le bruit numérique, mais jamais l'auteur n'indique si c'est à 100% écran ou pour la même taille de tirage, donc aucune conclusion à en tirer. 

[Jean-Claude]

Mais si il dit clairement comparer les images en crop 100% et moniteur 110 ppi pour ses évaluations de bruit :

« when viewing raw files at 100% magnification on a 110 ppi monitor). So... »

[JMS]

Mais si il dit clairement comparer les images en crop 100% et moniteur 110 ppi pour ses évaluations de bruit :

« when viewing raw files at 100% magnification on a 110 ppi monitor). So... »

Lu trop vite. Donc qu'l y ait plus de bruit sur un D850 que sur un D5 est une évidence, à 110 dpi il compare une image de 1,25 m de large à une image de 1,90 m...

[Fogger]

Toujours pas de traduction en Breton.
Seconde partie "Burst Depth II"
15 Octobre
http://www.naturalart.ca/voice/blog.html

Toujours aussi polyvalent ce boitier.

[big jim]

Lu trop vite. Donc qu'l y ait plus de bruit sur un D850 que sur un D5 est une évidence, à 110 dpi il compare une image de 1,25 m de large à une image de 1,90 m...

Brad Hill justifie son choix de regarder le bruit à 100% dans sa dernière mise à jour, en lien avec la tentation de croper fortement dans des fichiers de haute résolution.

[4mpx]

Brad Hill justifie son choix de regarder le bruit à 100% dans sa dernière mise à jour, en lien avec la tentation de croper fortement dans des fichiers de haute résolution.

Il ne le fait pas tout le temps.
Dans cette dernière mise à jour, il n'y a pas de comparaison mais il s'agît uniquement des crops sévères d'où la nécessité de vérifier la qualité du fichier d'origine à 100% avant de copper.